Good morning, welcome to tips and tricks for optimizing programming for cardiac implantable electric devices. It's good to see a solid crowd of people that realize the importance of device software and that there's more to the hardware just than implanting devices. And rest assured over the next hour you won't have to worry about hearing the phrase pulse field ablation. I'm Dr. Sunil Sinha from Johns Hopkins in Baltimore, Maryland. And I'm joined by Dr. Prakash, I will learn. I'm joined by Dr. Ritika Prakash at Dalhousie University in Halifax, Nova Scotia. And we'll start off with a presentation from Dr. Martin Stiles from the University of Auckland discussing optimal ICD programming by different manufacturers. Welcome everybody. First thing on a Sunday morning, well done. My name is Martin Stiles. I'm from New Zealand. I've the opportunity to talk to you about optimal ICD programming by manufacturer. These are my disclosures. So you might wonder why someone from New Zealand is talking to you about optimal ICD programming. Well, I had the opportunity to co-chair this 2015 expert consensus document on how to optimally program ICDs. And it was the first time that the all four continental heart rhythm societies, heart rhythm, APHRS, Latin American HRS and ERA got together to come up with the expert consensus document and there were four co-chairs of which I was one. And then four years later, we decided to update these 2015 expert consensus statements. And we refined them a little bit. Not a lot in the recommendations, but what we did do is we had manufacturer specific translation of ICD programming recommendations, which we hoped would provide a really practical way for people to optimally program their ICDs by speaking the language of the manufacturers and this was published in each of those four societies journals. So just to think a little bit of background about that 2019 update, there was a tension between studies and reports. So we had given therapy reduction programming because we knew that it resulted in a reduction in mortality. This is a meta-analysis, but even the meta-RIT trial showed a reduction in mortality with the therapy reduction. But on the other hand, there were a number of case reports published that said we had programmed the ICD in the way that you told us to and yet the patient died. And that was quite confronting for us because we had come out with these recommendations and yet you could report back in exquisite detail the demise of a patient. And so we updated these manufacturer specific recommendations to minimize such adverse events. But I think the most important thing was the patients who do not receive unnecessary ICD are not aware of being spared potential harm. Whereas patients in whom the ICD failed to treat life-threatening arrhythmias have their event recorded in exquisite detail. So it's much like when you prescribe an anticoagulant, the patient doesn't thank you for the stroke you prevented on Monday, but they blame you for the bleed that you caused on Tuesday. So these revisions employed the principle that randomized trials and large registry data should guide programming more than anecdotal evidence and that the recommendations should not replace the opinion of, obviously, the treating physician with a shared decision making outcome with the patient. So I want to just recap on the difference. There are inappropriate shocks and unnecessary shocks. We're all familiar with inappropriate shocks, shocks for AF generally or artifact, but there are unnecessary shocks. This is shocks the patient didn't need to have because the ventricular tachycardia was going to be non-sustained and terminate anyway, or it was going to be ATP responsive. So we had three things that we really said in these recommendations in 2015. Firstly, allow time. So for ICD patients, tachycardia arrhythmia detection duration criteria should be programmed to require the tachycardia to continue for 6 to 12 seconds at least, or 30 intervals before completing detection. And the aim of this was to reduce total therapies. And because of the large amount of evidence in the area, we were able to give this a class one level of evidence A recommendation. Secondly, we said only treat the fast stuff. So the slowest tachycardia therapy zone limit should be programmed between 185 and 200 with the aim of reducing total therapies, again, class 1A recommendation. And thirdly, leave SVT alone. So trust your discrimination algorithms to distinguish SVT from VT for rates faster than 200, even up to 230 beats per minute in order to reduce total therapies. Again, class 1 recommendation, slightly lower level of evidence. And so this is, and then we're able to translate these into manufacturer specific. I've chosen Medtronic, but it could have been any of the five major ICD manufacturers. We were able to give bradycardia programming recommendations for one, two, or three lead devices. We were able to give detection advice. We were able to give therapy advice, and we were able to give SVT discriminators, plus a number of other things at the bottom of the page. So I'm going to go through these in a little bit of detail, hopefully not too much excruciating detail. But for instance, this is an alphabetical order. This is the Abbott Medical. So what we said in patients in whom we didn't know what the VT cycle length was, we should program the VF zone to detect at least 30 intervals before calling it detected at a rate of 240, 250 beats per minute. For VT, over 187 beats per minute, and you could have a monitor zone if you wanted. If you did know what your VT cycle length, you could also add in a VT zone, which included that, and we suggested 10 or 20 beats below the known VT rate. And once you detected VT, you had a number of therapies. So VF, we had an ATP while charging, and then we told people how to program their ATP. Eight pulses at 85% VT cycle length, all shocks at maximum output, unless, of course, you'd done a DFT. And, of course, in this particular manufacturer, the first shock is a little lower than the subsequent shocks. And then for VT2, at least one burst of ATP, probably more, and we told people how to scan and that sort of thing. So this is the example in a little bit of detail for Abbott, but there's the same thing for Biotronic. So what are the differences? It's 30 out of 40 intervals. Some of their legacy devices couldn't go that long, so it's 24 out of 30. It was, again, 30 intervals. It was 188 beats per minute, but you'll notice the VF zone is 231. Why is that? Because Biotronic, the SVT discrimination algorithms are linked to therapy zones. So in order to safely put your therapy zone at 230, you couldn't have it at 250. You had to program it this way. There were other ways of doing it. You could use all three zones and you could actually treat 250 and above in VT, as long as you had the other two zones enabled below it. And again, if you knew that the VT cycle length, you could include that as well. And again, one shot of ATP for VF, eight pulses, 88%. You get the idea. Boston Scientific, this was a little bit different, of course, because in Boston Scientific, you had this fixed eight out of 10 detection intervals. So in order to delay therapy, and we said that was one of our aims, we had to put a delay on it. So for VF, and for Boston, you had two options. You could go the delayed therapy option, or you could go the high rate therapy option, much like they made at RIT trial. So the high rate therapy was very simple. You just only treated stuff over 200 beats per minute. In the delayed therapy, you were going over 185 beats per minute, and it was 12 seconds delay for rates between 185 and 250, but over 250, there was a five second delay. Why a five second delay? Well, what we worked out was if we wanted to apply 30 intervals to detect, we had to figure out how long that took. So it takes, at the rates we're talking about, 240, 260 beats per minute, it takes about seven seconds to detect 30 intervals. And so it takes two seconds to detect eight out of 10 intervals, so we had an extra five seconds to wait. So that was how we figured that out. And I guess it was an approximation because we were using studies from other manufacturers and translating them to Biotronic Ease, if you like, to try and provide this for people. And much like the other ones, you had VF therapy with maximum output shocks with a bit of ATP to start with, more ATP in the lower zones, and we had a coupling interval that you can read there for yourself with a 10 millisecond decrement. Medtronic, there was a relatively simple way to program Medtronic because of the way the device was set up and the fact that many of the trials came from Medtronic. So 30 out of 40 intervals over 188 beats per minute, and the VF VT thing just sorted itself out. But you could actually get quite complex with Medtronic if you wanted to. The other difference is that in the VT zone, we put 24 intervals. You can see if you knew what the VT cycle length was. If we were using a VT zone because it's a consecutive count, counting up, counting down, and this was data from the pain-free SST trial that guided us here. Again, therapy much the same as the other ones. Lastly, Microport CRM. This was a little bit similar to the Boston device. You had a fixed six out of eight majority, so we added extra 20 cycles, and you can see the three zones there, and they had quite an interesting therapy because they distinguished between stable therapy, in which case ATP is given, and unstable therapy, in which case they go straight to the shocks. So that's a guide to how we did detection and therapy for the five companies. SVT discriminators are as complex as you want to make them. There is onset, sudden onset, or chamber of onset, and interval stability. The thing about onset, though, it's a one-time discriminator. So you can't really base a lot of your withholding of shocks on that, perhaps. And there's dual-chamber algorithms, which you're familiar with, which compares the ventricular rate with the atrial rate and makes conclusions about that. And then increasingly important is these morphology algorithms, where they're matching the normal QRS template as seen by the ICD to what is seen in tachycardia. So this is just one. I'm not going to go through all these. You'll be pleased to know this is the most complex one. This is the Abbott one, where you could see we gave recommendations for single-chamber devices, which in this case was based on morphology. But you can see multi-chamber devices. You could also use this A versus V kind of algorithms along with it. So since 2019, when we published these, there's been a few papers. I'm just going to take you through just a few of these. So this is looking at how much ATP you can give. So this is for VT between 150 and 200 beats per minute, published in 2020. And you can see here, there is an incremental benefit from going from four, five, and six shocks. But it comes at a cost of accelerating the tachycardia. And you can also see that maximum output shocks are more effective than low output shocks. Since then, there's different types of ATP. You're familiar with burst. You're familiar with ramp. You're familiar with intrinsic ATP, which is really elegant, looking at the return cycle length and adjusting subsequent ATP intervals based on that return cycle length. And we know that intrinsic ATP has a higher VT termination rate and a lower acceleration rate compared to conventional ATP. And I'm really looking forward to the results of that. And lastly, do these optimal programming recommendations work? Well, I was really pleased to see this paper out of Thailand last year, looking at nearly 800 ICD patients followed for a couple of years. And you can see that if you're guideline concordant, you are getting less shocks than if you're non-guideline concordant. And the rate was about a 68% reduction if you had AF, and 44% even if you didn't have AF. And that was really gratifying to see. So just to conclude, this optimal ICD program by manufacturer, I recommend that you look at the 2019 focused update. This gives you specific therapy for a number of things. But the aim is to minimize shocks while maintaining safety from sudden cardiac death. Remembering that ATP is effective, painless, and energy efficient with newer algorithms. And lastly, shocks are still necessary and remain so, but should be minimized. Thank you. Thank you very much. I'd remind everyone to try to take a picture of that QR code so you can ask questions through the app. And I'd like to invite the next speaker, Dr. Prash Sanders from Adelaide, Australia. Thanks, Rathike. So I'll just let this go. They have definitely made me look like Vivek Reddy, but I assure you I'm not. So look, I've been asked to talk about reducing false positive in ILRs or ICMs and the programming side of things. And I've extended this a little bit because I think this is an important area to manage the whole patient because it's increasing the amount of alerts that we get in our remote monitoring clinics. These are my current disclosures. So there are a number of reasons why we are monitoring devices for ILRs. We use them for cryptogenic stroke. We use them for AF management. We use them for syncope. And this is a slide from Catherine O'Shea where she used the pacemate data of endoneutral system over a 12-month period, and 18% of these patients were loop recorders. And what you see on the figure on the left is that although the number of loop recorders is small, the number of transmissions that you get from this is enormous compared to the other devices. And what's more, in terms of the severity of these alerts, the red alerts are only a few, and you actually have a huge amount of yellow alerts that you need to deal with. So this is a significant burden to our clinics when we're dealing with remote monitoring. Now similarly, when we extend this to look at loop recorders, and this is done over 23 centres using the pacemate data again, based on the type of alert, and what you see is with asystole or ATAC and AF, you have a huge amount of false positive results, which means the positive predictive value of those alerts is quite low in our clinics. And so you're spending a lot of time dealing with an alert that doesn't mean so much to us. This has been reproduced in a single centre study by the group from Ohio, but interestingly they added to this that depending on your indications to have the device, you may actually have more false alerts. So if you're having cryptogenic stroke or syncope, you're more likely to have false alerts. It's been studied in the form of a systematic review, and there are 12 studies here, some retrospective, some retrospective. The largest of this series is in fact Dr O'Shea's data. But importantly in this column here, you see the burden of false alerts that occur, and it's very, very high as demonstrated in the previous slide. Now the problem may come from the start. Not all devices are equal, and in this systematic review, they included five different devices on the market, and you can see the different sensitivity, specificity, positive predictive value and negative predictive value. And you'll appreciate that in terms of positive predictive values, these are quite different. So the device you choose may influence the burden that you're creating in your clinic itself. I'll go through that a little bit more. In terms of the type of false results that you get, in this review, 51% was actually from ectopic beats. So you're talking about ventricular or atrioctopy, 19% due to noise, undersensing in 17%, and then you have a smattering of other things. And so now you start to work out what we may need to be able to program around. Let me show you an example, and this is from Sunit Mattel's paper. So all of these rhythms have been labeled as atrial fibrillation. The panel A is actually atrial fibrillation. Panel B is atrial flutter with some irregularity. C is atrioctopy, D is ventricular ectopy, E is where you have some noise introduced into it, and then F is T wave oversensing. So there are a number of different things, and how you manage this and program depends on what that false alert is on that individual patient. Let me take you back to this concept that you need to pick your device. This is work from Surya Kamsani that we presented last year at this meeting, again from pacemate data, equal numbers of loop recorders from four vendors, and you can see the type of alerts that you get and the actual positive predictive value, and you can appreciate that some are not very good at detecting ATAF, and some may be not so good at detecting syncope. If you look at the pool data, the blue shows the positive or true positive results, and so there's significant variation in the devices that we choose, depending on the indication and what you're using it for. Now the proviso here is that these devices are improving on a regular basis, and so each company does change as we go, and we need to keep ahead of this. Are there things we can do at implant? And this is a neat study. It looked at inadequate R-wave as defined by a change in amplitude of the R-wave during the follow-up, or noise being introduced due to artifact, and what they found is all of the false positives occurred in those patients. The only negative with this is that inadequate R-wave couldn't be predicted by any characteristic of the patient, except that if you put it in a non-parasternal location, you were more likely to have inadequate R-wave sensing. Now in the start of these devices, we used to undertake vector mapping to find the biggest amplitude signal, and we've given up on this because the concept was you can put it anywhere in the front of the chest and you'll be fine, but here's a neat study that's currently just come out in HeartRhythm. They did undertake vector mapping. They were able to look at getting a better amplitude of the electrogram, and in fact the better electrogram amplitude reduced the false positive results. This may be significant. Maybe we do need to go back to mapping this and implanting to reduce our workload in the clinic afterwards. We've kind of assumed it's not such an issue. Now once you've implanted these devices, it's important for us to review the alerts so that we can program this appropriately. This is data that's come from Europe. It looked at the Biomonitor 3, and it's very interesting. They had those patients who had more than 10 false positive alerts in the first month. This is enormous, and they had a protocol-based reprogramming of whether it's changing the sensitivity to avoid the ectopic beats, whether it's increasing the number of counts to detect a tachycardia so that it can reduce the burden, or increasing the sensitivity of the AF sensitivity. But what they did find, that although a small number of patients, they were able to reduce dramatically, 98% reduction in this false positive result. Now that seems a little bit excessive to wait to 10 false positives, and something that we should consider more regularly. This is again from the group from Ohio, and they had a nurse clinic that brought these patients in and reprogrammed the device and were able to significantly reduce the burden, and not surprisingly, a huge improvement in cost into their clinic of undertaking this sort of programming. What can we do? We can tailor based on the arrhythmia that you want. You have device-specific settings, and all of the manufacturers are making this rather automated for us, so there's not a whole lot of variation that you need to start with. You can adjust detection duration, sensitivities, and also the blanking period, and the blanking period's a little bit more complicated depending on the device that you use. So this is from the LINC data, and there are so many programmable elements to this. How much of a threshold you have, how long you start before it starts decaying, and when it drops to detect, do not lose information. Each device has slightly different programmability, but you can go into this if you wanted to. Now having said all that, we're moving into the area of AI doing most of this for us, and I think we're gonna see a lot more of this coming out. This is from Sunit Mattel, and they used the LINC device with cardiologs in Paris to look at not just the ventricular electrogram, but also the atrial signals to put it through a neural network to give us an answer of whether they had AF or not. Three different indications, but look at the marked reduction in green in the false positive results that they got from this. So I think this is going to be a reality, and the fact is that Medtronic have taken this and introduced something very similar in the back end of their CareLINC system, so it's already cleaned for you before it comes to the clinic. This is work from Melissa Middledop for using, again, the Pacemate data. It's in press at Heart Rhythm. So before the AcuRhythm AI algorithms were introduced, and after the, sorry, before and after, and green is after, and what they showed was that there was a 20% reduction in alerts per patient and a 40% reduction in the episodes per patient, so there is an improvement or reduction in false positive results that we can get. Now, this is getting even better for us because now we can remotely program devices, so combining the AI with remote programming, we may actually be able to do this without having to bring the patient to clinic. Two devices do this, the LINC2 and also the LUX device, and this is a very neat study which shows in the LINC from LINC1 down to the LINC2 and then down to what we get with the AI algorithms and reprogramming. Similarly, with the LUX device, you can see the number of programmings that needed to be undertaken here, 18 and 37 in this small group of patients, but look at the marked reduction in the median, the alerts per month that they got as a result of doing that, so we can achieve, by reprogramming, significant reduction in this. There are other ways to do this, and here's a small series that Melissa led. This is using a pilot study of the pacemate managing this for the clinic. A small group of patients, only a small number of ILRs in this, but what it does show is in standard care where the clinic was managing it, 55% had yellow or red alerts, and this reduces down to 22.7% if you manage it with intensive management where there is a programming that's done. Again, here, total alerts, significantly better when they were intensively managed, so there may be an opportunity to use proprietary software to be able to do this. Now, having said all that, it's really important for our clinics to have management pathways for how you deal with clinical things, and this is a concerning thing, again, from Cat O'Shea, including all devices, including loop recorders, but what it shows is this group of patients who, for loop recorders, would have had a CHA2DS2-VASc score of over two and AF detected for over six hours or over 24 hours. A significant proportion at 12 months are still not anticoagulated, and so we may be good in detecting. We may try to reduce false positive, but we gotta do better here, so let me conclude at this point. How can we avoid this? We can pick our devices, we can optimize the position, we can start with nominals, but we need to review the programming, and here's the flow diagram from the University of Ohio. These are the programming elements that we can use in order to reduce our clinic burden. Thank you. Thank you, Dr. Sanders, and now we invite Dr. Brett Atwater to the podium. He's from Inova Heart and Vascular, in beautiful Fairfax, Virginia, and he's gonna speak to us about optimizing programming for physiologic pacing. Great, thanks very much for the opportunity to come and speak. These are my disclosures. Still working to get parity with Vivek Reddy on that as well, so programming tips to improve physiologic pacing outcomes. There are some assumptions for this talk, and the primary assumption for this is that all devices are bivy pacemakers or left bundle pacemakers, which have LV septal or left bundle capture, or his bundle pacemakers, as defined in the 2023 consensus statement, and all these devices have Bachman's bundle atrial leads. Everything else is really not capable of physiologic pacing, and so different programming is requested and required for those, primarily to try to avoid atrial and ventricular pacing and the dyssynchrony that those can induce to the atrium and the ventricle. Currently, we have pacing indications for AV block, for ventricular resynchronization therapy, or CRT, and for sinus node dysfunction, and then you are probably all aware after this meeting of emerging pacing indications for HEF-PEF with relative bradycardia and atrial resynchronization therapy. I'm not gonna go into those with much depth, simply because there's not a lot of data about how best to program devices for those indications, so we're gonna focus primarily on the current pacing indications. So really, your programming comes down to an understanding of what are your goals for the pacemaker, what are you trying to achieve with the pacemaker, and for AV block and ventricular resynchronization therapy, your primary goal is to provide safe ventricular rate support, so that's the number one priority, is to make sure that the patient receives pacing when they need it. Second goal is then to prevent AFib and atrial, and heart failure by providing a synchronous ventricular contraction. In biventricular devices, that's by promoting RV and LV capture and optimizing the timing of each of those, and then with left bundle and his bundle pacing, it's to promote capture of the conduction system, not just the septum and not just the myocardium. Our third goal then is to promote device longevity by minimizing unnecessary pacing and managing device outputs. For sinus node dysfunction indications, we're trying to provide atrial ventricular rate support and prevent heart failure and AFib by providing synchronous atrial contraction and minimizing ventricular pacing, and then promoting device longevity again by minimizing unnecessary ventricular pacing and managing our device outputs. So let's start with programming in a patient who's got an AV block or a CRT indication. The lower rate for those patients who have these devices who have HEF-PEF should probably be according to the MyPace lower rate, and this we'll go into a little bit more in a following slide. For patients without heart failure or for patients with HEF-REF, there's a lot of debate about what that lower rate should be. Should we be using MyPace for that indication? Should we be using lower rates? Or should we be using the out-of-box setting of 60 beats per minute for most devices? And at this point, I think there's not a lot of data to understand how to best do that, and so most people are leaving that at the out-of-box setting of 60 beats per minute, but more to come in hopefully next year's HRS. AV intervals, for those patients who have Bachman's bundle leads, we like to pace the AV intervals with a pace and sense AV interval being identical. And the reason for that, why we don't need to have a longer paced AV delay is because Bachman's bundle pacing does not introduce interatrial conduction delay. And the reason why we're pacing 30 milliseconds or more prolongation of the AV delay with paced AV delay is to make up for the interatrial conduction delay that pacing produces compared to sinus rhythm. So if we're not doing that anymore, we don't need to give you that extra 30 or 40 milliseconds on the paced AV delay side. So I like to program mine identical because there's no difference in left atrial activation time with pacing versus sinus rhythm. For AV interval and CRT indication, we are really primarily trying to use the AV interval to promote ventricular synchrony. And so we program this at the time of implant to really minimize the QRS duration or QRS area if you have the ability to measure that in real time in the EP lab, which I'm still struggling to get, but would love to have. And then what I do is I find the AV interval that produces the shortest QRS duration. I program the rate adaptive AV delay on in order to promote that at various levels of atrial rate. And then if we're using a Bi-V device, I like to use the adaptive CRT feature if I'm using a Medtronic device in order to always try to promote fusion of left ventricular pacing and right bundle activation. In general, for AV block or CRT indication, I like to program rate response off unless they also have a sinus node dysfunction indication. The reason is because there's compelling data that unnecessary atrial pacing in CRT-indicated patients actually results in worsening outcomes. And it's unclear whether that's also true in a Bachman's bundle position, but clearly true in a right atrial appendage location. So unless you really need rate support and you have a lead that's not in Bachman's bundle, I'd like to turn rate response off. Programming in a patient with AV block continued. The AV block P-wave duration plus about 30 milliseconds is useful to try to prevent A-wave truncation. That's what you're seeing on this section of the slide on the right here. This is your E-wave. I don't know if you can see my pointer or not. No, you cannot. You have an... Swing the cursor to the big screen. We'll be able to see you. There we go. Okay, thank you. So you've got an E-wave on the echo and then you've got an A-wave here and you can see the terminal portion of the A-wave gets cut off at the same time that you've got ventricular systole. And that's happening because of the fact that your interatrial conduction delay is actually producing more prolongation than your AV delay. So the ventricle fires before the left atrium is finished and as a result, you get A-wave truncation. This results in really high left atrial pressures and we'd like to avoid this. Way to do that simply without having an echo at the time is to look at your P-wave duration, add about 30 milliseconds to allow the P-wave to finish before you fire the ventricle and then fire the ventricle with your AV delay there. There's a lot of debate about unipolar versus bipolar programming for left bundle area pacemakers. There are advantages to unipolar, the most important being you can assess the presence or absence of an R-prime to understand whether you're truly pacing the left bundle area or are you pacing the RV septum. You don't have that confusion that could be created by anodal capture in the setting of bipolar pacing. Some disadvantages to that though is that currently it would require reprogramming if the patient needs to get an MRI, which can be a lot of work for you or your rep or whoever you have working to do that. And so really we choose whether to use unipolar or bipolar based on expected battery longevity and whether we are trying to promote or avoid anodal capture. It's still unclear whether anodal RV stimulation from bipolar pacing is good or bad. It probably depends on the patient's underlying physiology, whether they have complete heart block or whether they have left bundle and where you can still promote fusion with right bundle conduction. Programming in a patient with a sinus node identification who has a narrow P wave to begin with, for the lower rate in a setting of HF-PEF, we would likely wanna program that according to the MyPace lower rate versus 60 beats per minute. So a little bit unclear, retrospective study at this point, prospective study at this point, but only one study which was single center, little more data is gonna be necessary. It will be forthcoming in the coming years. So a little unclear whether to use MyPace or 60 beats a minute. I prefer the MyPace lower rate. This is how to calculate the MyPace lower rate. It's a nomogram which takes into account the patient's ejection fraction and their height and then you can calculate what their lower rate should be for that and program it accordingly. There are differences for men and women. For the AV interval in a patient who's got a sinus node dysfunction indication, if they have a PR interval, I'm sorry, I program it AIR, DDDR, or long AV delays in order to maximize battery longevity if they have a pretty short PR interval. But if they have a long PR interval, and again, we have a conduction system leading the ventricle and they have HEF-PEF, we like to program AIR, DDDR with a maximum AV interval set to 250 milliseconds in order to maintain AV synchrony. So really long AV intervals can result in other performance issues for the ventricle and so we like to try to keep that PR a little bit shorter in that case. For sinus node dysfunction in a narrow P wave, we turn rate response on for those who have chronotropic incompetence, but off probably for those patients who just have resting bradycardia, but are still capable of mounting an appropriate heart rate with exercise. And that's sort of two different types of sinus node dysfunction that we see commonly. They need two different programming methods for that. For patients who have sinus node dysfunction in a wide P wave, the goal there starts to become, let's try to narrow that P wave and get better left atrial mechanics and potentially prevent atrial fibrillation and HEF-PEF symptoms. So in this group, for HEF-PEF, we definitely wanna use the MyPace lower rate and we're trying to promote atrial pacing so that we can promote atrial resynchronization in these folks. For those who don't have HEF-PEF, we may use 60 or we may use MyPace. We're trying to get to the one that's gonna promote a high atrial pacing burden. We turn atrial preference pacing on in these patients. This is an algorithm that allows you to always be about five beats per minute above sinus and that allows you to constantly be overdriving the sinus node and producing atrial resynchronization therapy. For these patients, similar AV recommendations. If you have a narrow PR interval, then we try to avoid pacing the ventricle if possible. If you have a really long PR interval, then we may try to resynchronize AV intervals in order to promote AV synchrony. And then again, rate response we turn on, in this case, in order to try to overdrive that sinus rate during exercise and that allows us to try to synchronize the atrium both at rest and with exercise. Gonna dive into atrial ATP algorithms, which is on the Medtronic platform, the reactive ATP. Should we be using this commonly in every patient? The Minerva trial is the best available data we have for this. They compared DDDR mode to MVP mode and then DDDR also to MVP with reactive ATP turned on. And these are the event curves here and you can see the blue line here is the effect of turning reactive ATP on. It pretty significantly reduces likelihood of short and long duration atrial fibrillation. Also reduces the probability you're gonna need a cardioversion. It's currently off when it comes out of the box and there's a recommendation that you turn it on once you get out to the first clinic visit. Turns out that when we look at CareLink, it's hardly ever turned on in clinic and so I like it on, so I program it on at the time that I do the implantation. That is off-label use, but there is a device algorithm built in. The reason why we don't do that is because if the atrial lead falls into the ventricle, you could inadvertently deliver ATP to the ventricle, which can initiate VT, bad, in a patient with a pacemaker. But there is an algorithm built into the device that can detect that the lead has fallen and automatically turn off ATP. It works in greater than 99.9% of cases of atrial lead dislodgement, so it really becomes a non-issue as long as you're using a modern-day device in these patients. Important considerations for HisPundle. There are a lot of important considerations in HisPundle. This is an entire talk about how to handle programming for that, but some of the most important issues just to address quickly are, in patients with AV block, again, the first goal of the pacemaker is to produce ventricular rate support. Programming the output to a non-selective capture as long as it's less than three volts per minute allows you to make sure that if you lose HisPundle capture because of a rising threshold, you'll still have ventricular rate support by pacing the ventricle. So if you have the option of selective capture at one volt or non-selective capture at two volts, I like to program to always try to promote non-selective capture to make sure in a patient with heart block that I'm not gonna leave them with asystole because I'm programming the output too low. If the His lead is in the RV port, the sensing vector, bipolar, unless the R wave is less than two, or if you have atrial and His over-sensing, then we try unipolar to get away from over-sensing problems. Ensure that there's no over-sensing of the A or His potentials. Consider automatic sensing thresholds if the R wave is small in those cases. If you are over-sensing in both unipolar and bipolar, really have to perform a lead revision because you've gotta be able to sense and prevent inappropriate inhibition in a patient who's got AV block. If the His lead is in an LV port, we wanna try to promote pacing off of that lead instead of off of the RV lead, which may be in the other port. And so we always pre-excite that LV port maximally to try to get as much His capture as we can before that RV lead paces unnecessarily. If RV backup is not needed, we like to turn off or turn down the energy on the RV port as much as possible. You can't turn it all the way off on most devices, but you can turn the energy down as much as possible in order to reduce battery drain for that. For ventricular sensory response, you have to turn that off if you have a His lead in. It doesn't work, and it's gonna drain the battery unnecessarily. And then for unipolar and bipolar pacing, capture threshold's usually lower in unipolar, but the impedance is also lower. So really choose this based on the impact on battery longevity, not on the threshold itself. Can I use capture management with CSP? You cannot. If the CSP lead is in the atrial port, you can. If it's in the RV or LV port, you can't. To do that, though, you really have to make some fine adjustments. Those fine adjustments can be seen in this case. This is a 55-year-old woman with heart block. Her rate was 25 beats a minute with an EF of 45%. We put a left bundle pacemaker in. The goal was to get LV septal and left bundle pacing, again, because she has heart block. Her LV left bundle threshold was 2.5. Her LV septal threshold was one. And so after the acute phase, what would happen here is that we would drop the output with capture management to two volts, and that would then put us below the left bundle threshold. So we would be doing LV septal pacing instead of left bundle pacing. So what you wanna do in this case is you can actually bump up this minimum adapted amplitude, or you can bump up the amplitude margin so that you will always have capture management running, but you will always be above the left bundle capture threshold. And then this is an example of capture management with CSP, so something to avoid in older generation devices. If you're running atrial capture management, the way that the devices work are that they will deliver backup pacing through the RV port. And if you have done what I just told you to do, which is turn the output down on the RV port to save battery and you have a patient with heart block, then when atrial capture management is running, there will be no pacing and no support in the ventricle because the LV lead gets turned off. So in these older devices, make sure that you turn off atrial capture management. On newer devices, it will pace by V, so you don't have to worry about it. They've fixed it, but on the older devices, this can cause syncope and end problems for the patient. That's all I've got. Thank you very much. Thank you. Thank you very much, and we'll invite, last but not least, Zach Winnett from the UK to speak. Thank you very much, and good morning. So I've got a slightly shorter disclosure slide. It's disappeared, here we go. So I'm gonna talk about pacing in delayed AV conduction. So if we take a step back and think about the main pacing indications, the reason that we pick patients to deliver pacing is that patients have developed a conduction, a form of conduction disease which leads to cardiac output. So in the first, the reason we developed pacemakers originally was for bradycardia indications cause profound reduction in cardiac output, causes symptoms for patients. And we ultimately, what we try to do is deliver pacing in as physiological a form as possible to try and restore normal physiological activation and thereby improve cardiac output and improve symptoms. Now having a long PR interval is associated with adverse outcomes. So here is some data from a relatively unselected population and we can see that if you've got a longer PR interval then you have a slightly higher risk of having heart failure events and possibly even atrial fibrillation and mortality. Heart failure patients in particular have more to lose. They've got less reserve and this is data from the ICD arm of the MADET study and you can, in patients without left bundle branch block, and you can see that those that had a very long PR interval had a greater risk of heart failure hospitalization or death. Now this of course, so having a long PR interval could of course just be a marker of risk. It doesn't necessarily have to be. It doesn't necessarily mean that it's a treatment target. So the patients could just be sicker but having a long PR interval does adversely affect cardiac function. So you can have, if you have a long AV delay, you get EA fusion which shortens the diastolic filling time. You get diastolic MR which decreases preload and so both these mechanisms can lead to reduced cardiac output. And of course it's something that we can treat with a pacemaker and this very nice study from the Maastricht group used computational modeling, animal studies, and some patient data to look at whether if we intervene on patients with a long PR interval, whether we can improve cardiac function. And they showed that if you shorten the AV delay, that it reduces the, improves filling and ultimately leads to improvements in stroke volume. So acutely we can see improvements in stroke volume. So maybe this suggests that a long AV delay could be a treatment target. Now biventricular pacing in patients with left bundle branch block essentially was the first successful therapy with pacing which was a non-bradycardia indication and was shown to reduce mortality and improve symptoms. And this was by not slowing the heart rate but improving the way that the heart is activated. And if we think about the mechanism of benefit with biventricular pacing, we mainly think about the disadvantage of being in left bundle branch block is to do with less synchronous activation and therefore less effective pumping of the heart. But of course if you develop left bundle branch block, you also can get prolongation of the left-sided AV delay and therefore adversely affect filling in that way. So there's two potential mechanisms through which cardiac output can be reduced. We tell our patients, and we're gonna give them CRT, that what we're gonna do is we're gonna improve ventricular resynchronization. But what's the contribution of AV delay, of AV shortening? So we looked at this in an acute study where we, within the same patient delivered biventricular pacing and his pacing, but we deliberately didn't correct the left bundle branch block. So we paced the his bundle with a left bundle branch block pattern and we looked at acute hemodynamic response. And what you can see is that without delivering any ventricular resynchronization, we get a significant improvement in acute hemodynamic function. And actually ventricular resynchronization contributes much less to the acute improvement. So AV filling may be a very important mechanism through which we improve cardiac output with biventricular pacing. So how should we program the AV delay in patients with CRT devices? So now we have more options with CRT devices. We can deliver biventricular pacing or we can do conduction system pacing. And what's often recommended in patients with left bundle pacing is that we program an AV delay that delivers the shortest QRS because it makes sense. If you think it's all about ventricular resynchronization, we deliver an AV delay that's longer and allows fusion so that we get activation via the intrinsic conduction via the intact right bundle branch block, and right bundle, and thereby deliver more rapid ventricular activation. And similarly with biventricular pacing, we can utilize fusion and get narrower QRS. But is this the best way to program the devices? Because I've just told you that actually filling may be important as well. So Dr. Liang looked at this and we performed acute hemodynamic measurements. And if you look here, this is systolic blood pressure. And in red is the AV delay that delivered the shortest QRS duration. Whereas here we can see that if you prioritize filling, overall you get better hemodynamic improvements. So the problem with the longer AV delay is it's further along on the AV curve that we see and actually it may not be best to always program the narrowest QRS. And what about pacing patients with prolonged AV delay who don't have left bundle branch blocks? So as a primary pacing target? Well, we know that it's probably in patients with impaired function and have less ability to tolerate the adverse effects of having a long PR interval. We can correct this with pacing and with conduction system pacing, we can potentially deliver pacing without inducing more harmful effects. The problem with RV pacing is you can improve ventricular filling but actually you induce ventricular dyssynchrony. So we looked at this in the HOPE-HF trial, which is where we recruited patients with heart failure, reduced ejection fraction and prolonged PR interval without left bundle branch block. And we performed AV-optimized his bundle pacing using acute hemodynamic measurements. So we used non-invasive beat-by-beat blood pressure and identified the hemodynamic optimal AV delay. And our primary outcome was exercise capacity and secondary outcome was blinded quality of life. And if we look at the overall results, we didn't have a significant improvement in the primary outcome of peak VO2, although there's a trend towards improvement. But patient-blinded quality of life was significantly better in those that received the pacing period, so when they had pacing on. So this was a crossover study, so they had six months of pacing on, six months of pacing off. And when we asked patients at the end of the study, so these are patients with heart failure, and we said, which period did you prefer? When did you feel better? And there was a strong preference for the pacing period. So they felt better when they had the pacing on. Now, of course, it's important in this pacing, in this group of patients with narrow cuirass and impaired ventricular function to find out what we did to the ventricles, because most pacing, even biventricular pacing in this group of patients, can induce ventricular dyssynchrony and cause problems. Well, there was no signal for harm in terms of ventricular function. So it seems to be, if we use a conduction system pacing approach in this patient group, that we don't harm their ventricular function. And then we've performed a sub-analysis, and we looked at the acute improvement in blood pressure at the beginning of the study, so where they had their optimization. And what we found was that if you had a positive response in acute hemodynamics, this was a very strong predictor of longer-term outcomes. So if you improved acute hemodynamics, then those patients got improvements in peak VO2, improvements in symptoms. And you may say, well, this is symptomatic improvement. What about improvements in mortality? Well, if you ask patients what matters to them, so you have a patient who has an impaired ventricle and a narrow QRS, and you say you have an ICD indication, they often say to you, is it gonna make me feel better? But if you put an ICD in, the answer is, of course, no. So it's gonna save your life. And if you ask patients, like the BHF did in this heart failures, in this survey, they found that three of the top five challenges for patients were symptoms. So this really does matter for patients. We don't necessarily need to make them live longer. We may actually, having a treatment that can make them feel better can be really important. What about patients who have normal function? Well, there is actually guidance, and it's included in the guidelines, that if you have a very long PR interval, then you can offer patients a pacemaker. But you have to really clearly show that the symptoms are attributable to having the long AV delay. And certainly, there's some patients where it's clear. Others, it's a little bit more difficult to work out. And sometimes, the only way to know is if you pace them and see if they feel better. So in conclusion, delayed AV conduction can lead to reduced cardiac output. Shortening AV delay can lead to acute improvements in stroke volume. And this improved AV filling seems to be an important mechanism in the way we improve cardiac output with CRT. Perhaps we should prioritize filling rather than the shortest cure us duration when we program our CRT devices. And isolated AV conduction is a promising treatment target for patients with impaired LV function. I accept that perhaps we do need to repeat the studies with symptoms of the primary outcome because there's always a little bit of concern about analyzing data, which is a secondary outcome. Thank you very much. Thank you. Thank you. Thank you very much to the speakers for fabulous talks. We have a lot of questions here online. So I'm going to start with asking some of them. The first one is for Dr. Atwater. Is there any concern for enabling HLA-ATP for patients not on anticoagulation? Yeah, so this is a really important point. And reactive ATP usually turns on within the first minute of atrial fibrillation. So if you're a believer that six minutes or more of atrial fib is required to be able to form a thrombus, then you should be okay. But there are very few data in this area. So we usually do not turn it on in patients without anticoagulation unless they don't have a prior history of a-fib. And then the first time it's activated, we anticoagulate the patient. So that's how we handle it. Yeah, that sounds good. I'm just gonna take the chair's prerogative and ask the question myself. How many on the panel are putting Bachman-Spundell atrial leads for every patient? Okay, one. And how about in the audience? Are you using Bachman-Spundell leads for atrial pacing? No, not really. Okay, so I was just wanting to know what that is sort of sitting like. Yesterday's Bachman-Spundell session was packed. Yeah. Is that right? Really, it was this room and the standing room only was amazing. Yeah, yeah. So it's something that we probably need to be considering. Yes, it's the only form of physiologic atrial pacing that we have available to us. It seems to have more promising outcomes than right atrial appendage pacing and even historical trials of biatrial pacing. So it's actually a very simple implant and I'm willing to bet it's gonna be a safer implant than right atrial appendage. So even if you don't get atrial resynchronization, the probability of perforation and lead dislodgement goes down by putting it there. So if you haven't tried it, work on it. The first couple even are pretty easy to do and then you end up with a much better long-term outcome, I think, and short-term outcome for your patient. Yeah, no, I think it's probably the way we have to go. We need better tools, I think, for it. And I don't know if the criteria for Bachman's bundle pacing has actually been well established as the only other consideration, yeah. That's the big thing at the moment. I mean, it's not difficult to put the lead in that position but we currently have challenges in knowing whether we've actually captured Bachman's bundle or not. But I accept it's not gonna be any worse than right atrial pacing. So you may get a narrow P wave just because you're pacing from the middle of the heart. Right, right. As long as the sensing's okay and the lead stays in and the capture threshold's not too high, then it's probably not gonna cause any harm at least. Yeah, and it's important to know, I think, too, that it's not like left bundle pacing. It's not an insulated sheath of specialized conduction fiber that you're trying to capture. It's muscle. It just happens to be muscle that is oriented all in the same direction and therefore conducts a little bit faster than other muscle. So it's not like you're trying to get a Bachman's bundle potential and then it's a very different process than what you're used to with other conduction system pacing and really it comes down to your P wave and can you narrow the P wave compared to sinus rhythm? You're almost always gonna narrow it compared to right atrial appendage pacing, which is what Dr. Winnett was getting at. It's probably almost always better than that but can you do better than sinus? That's where the criteria really need to be developed. Okay, great. So there's a few common questions here. Dr. Winnett, do you AV optimize patients in the cath lab or clinic and do you do an echo for every pacemaker implant to set AV delays to maximize filling? Yeah, so it's a good question. So there are different ways you can optimize AV delay in terms of using, what are you gonna use to tell you how to program the device? For me, the best thing is to have something that's the common endpoint because there are various different things that you can prove filling, you can prove ventricular resynchronization and patients may have a longer AV delay. So there are various different things that come into play. So for me, the best marker is something that's a common endpoint. What we're trying to do is improve cardiac output. So we use noninvasive blood pressure beat by beat. We don't do repeated optimizations because it seems to be quite relatively stable, although the best thing would be for the device to have a hemodynamic way of doing this and doing it repeatedly, doing exercise and things. But for us, it's noninvasive blood pressure, except that's quite difficult if you don't have the setup. So echo optimization is also fine. Maybe if you use cardiac output, but you do need to use repeated measurements and averaging. So don't just use the average of one beat because there's a bit of variability. And we probably don't use echo enough in our patients after pacemaker implants. We don't look for TR. We don't look for that AV optimization. So it's probably something that needs to be improved. So Dr. Sanders, there's a question here on how do we detect syncope from the remote monitor system? How do you detect syncope? Good question. So, I mean, basically the syncope is bradycardia is what you're kind of looking for in terms of an alert. There are algorithms in most of the loop recorders that do tell you position. It just hasn't been turned on yet. So we may have that coming out in the not so distant future. Yeah, that'll be very helpful. So last question probably for Dr. Stiles. How do you use morphology match algorithms in complete heart block where there's no intrinsic QRS? Yeah, you can't. Yeah, you can't. And this is one of the, I gave the talk in the very briefest sessions. There's a lot of asterisks and things like, for instance, obvious stuff like SVT discriminators in complete heart block, turn them off. And the other things that, and one of them was not, you can't do the match. The other thing is you have to turn off the auto update. So there's a lot of tricks, and I was a bit like your talk with, there's tricks, isn't there? There's fishhooks and everything. So yeah, the devil's in the detail, unfortunately. So you cannot use those, yeah. And can you comment on syncope with shock reduction programming? Because patients with non-ischemic cardiomyopathy with a very poor pump often have syncope when you delay. Yeah. Yeah. And so certainly there is a small cost of syncope. In the trials that were done when you put longer detection intervals, you can get a small amount of syncope, but surprisingly little. I think it was a real fear for people. You remember when it was 12 out of 16 or something? These things were set to fire. I don't think our concerns about syncope were borne out. So it is a risk, and there are some patients who do have syncope with longer detection intervals. There's also some patients who really hate the shocks and want to be unconscious. I can't understand that, but they do, don't they? Yeah, for sure, for sure. That's always a concern, though, because sometimes, or even in ischemic, where detection's delayed, they get more ischemic as they have ventricular arrhythmia, and then you, but that is rare as well, so. All right, I think that's time. Thank you, everybody, for your questions, and for the speakers, my chair. Thank you.

cardiac device optimization

pacing strategies

ICD programming

implantable loop recorders

physiological pacing

AV conduction

false positives reduction

programming recommendations

arrhythmia treatment