false
Catalog
Inside EP: Insights From Clinical Decision-Makers ...
Device Therapy: Therapy Programming and Unmet Nee ...
Device Therapy: Therapy Programming and Unmet Needs
Back to course
[Please upgrade your browser to play this video content]
Video Transcription
The goals of defibrillator therapy, as you all know, is to terminate ventricular tachyarrhythmia. This is gonna be a very defibrillator 101 at the beginning, hopefully more interesting towards the end for people who are less familiar. So this is just a picture of someone who's had a myocardial infarction, and you can see the scar and this kind of, you know, this tissue here versus the healthy kind of red tissue here. And what you get is sort of like these islands of dense scar, and then in between this more complex tissue that will still conduct, but a little bit slower, and then you can, this is the perfect setup for one of those reentrant circuits that Dr. Russo and Dr. Shivkumar were talking about, and a perfect setup for ventricular tachycardia. And then the other type of tachyarrhythmia we see is ventricular fibrillation, which is more complicated, much more disorganized, rapid rhythm that we wanna stop. So how do defibrillators work? Basically, there are two types of therapies that we can apply, either fast pacing, or ATP, or anti-tachycardia pacing, or just a good old, you know, pop to the chest with a shock. So how does that work? Well, I think we're all familiar with the, you know, external defibrillators, with the electrodes that go on the chest, and the idea is that these electrodes form a vector through which, you know, a high-energy therapy is delivered, and the goal is to deliver maximum current density through the heart muscle. That's the key to a successful shock. And so we've kind of recapitulated that with implantable devices. We can either deliver fast pacing with a pacemaker part of the lead, or we can use the shocking coil and the can now, as our electrodes, to deliver that high-energy shock. This is a transvenous defibrillator, with the wires going through the veins and into the heart itself. And we also now have the option of the subcutaneous defibrillator, where the can is out here to the side, and then you can see sort of just barely the lead coming across the chest, over the ribs, actually, so it's subcutaneous. It's no longer in the veins or in the heart, and then the lead itself is sitting on top of the sternum, so over the bone and under the skin. And again, the coil here and the can function as the electrodes to deliver that high-energy shock. So in terms of programming, there's programming that we can do to help the device make decisions about when to deliver a shock, when not to deliver a shock. So the very first parameter that we will decide about is the cycle length, or the beats per minute. And based on sort of fine-tuning over time, we've come to kind of 200 beats a minute as a decent starting point for people, especially for primary prevention. And the goal, on the one hand, is to make sure that we shock people if they're really having, you know, a tachyarrhythmia that's life-threatening. On the other hand, we wanna avoid what's called inappropriate shocks, and those rates used to be, you know, kinda high in the double digits for sure, and over time, as we've refined our thinking and sort of testing of patients, we've been able to drive the rate of inappropriate shock down to kind of the single digits. For younger patients who can get their heart rates very high just by, you know, running really fast, we're gonna set that cutoff rate higher. For patients who've had proven ventricular tachycardia, like Dr. Russo's patient, who had ventricular tachycardia at like 150 beats per minute, we're gonna program that rate lower so that patient gets treated for their individual slow tachycardia. The other thing we program is duration. So, you know, we, as, you know, I guess as the professionals in the industry, you know, kind of decided on lower, shorter durations at the start, but again, over time, we sort of realized that patients aren't at more danger if we extend that duration out more, because we realized that some of these tachyarrhythmias are gonna stop on their own, and so why deliver a shock, which we know causes significant harm. And so we've pushed the detection out to like 30 out of 40 beats, or two and a half seconds from sort of shorter durations in the past. And then another sort of criterion that helps us determine whether something should be shocked or not is something called waveform matching, where we take a template of what the QRS might look like in sinus rhythm, and compare that to when the patient has an arrhythmia. And if it looks the same, then that suggests it's actually probably more likely to be a supraventricular tachyarrhythmia using the same natural electrical system versus a VT coming from the bottom chambers of the heart. So here's an example, fresh, hot off the presses, of a patient who had non-sustained ventricular tachycardia. So because the rate was so rapid, so shorter than 300 milliseconds, it was appropriately tagged as being fibrillation or, you know, a tachyarrhythmia, but then it ultimately stopped, and because the duration was longer than this, the patient did not receive a therapy, which is a good thing. There was waveform matching performed as well, and you can see that the dotted line is the template. The device compares the, you know, fast heartbeats, shows that there's no match with the template, and appropriately characterizes this as, you know, ventricular tachyarrhythmias. So this is just, you know, I think this is just one, you know, sort of tip of the iceberg way in which, you know, a device can sort of automatically make these categorizations for us. So once we've gone through detection and a tachyarrhythmia is detected, what can we do? So we can deliver anti-tachycardia pacing. So how does that work? Well, Dr. Shivkumar showed you this beautiful sort of concept of reentry, where there's, you know, a fixed circuit, and the electrical signal is going around that fixed circuit. We can kind of look at it in a more schematized way, as Dr. Shiv showed, where sort of this leading edge of depolarization, so all of this tissue is refractory because it's just been activated, and then the tissue kind of behind that is somewhat refractory, and then finally there's this area called an excitable gap, where the tissue is completely recovered, and that's the only way in which the circuit can perpetuate, right? There has to be tissue in front of this leading edge that's allowing for continued perpetuation of electrical activation. And the idea behind anti-tachycardia pacing, which again is delivered by, you know, the pacemaker part of our lead, is that a pacing signal enters that excitable gap and then excites all this tissue, depolarizes it, and then that leading edge meets, you know, depolarized refractory tissue, and then just extinguishes the reentry circuit. Does that kind of make sense? And so, that's pretty cool. So, and how do we program that? We choose the percent cycle length of the tachycardia, so if this is like 260 milliseconds, we have to pace a little faster than that. We have to decide how many beats to deliver, because maybe we deliver a pulse and it hits, you know, tissue that's already refractory and it's not gonna stop the circuit. How many sequences will we deliver? Well, this was kind of, you know, first studied in the pain-free trials, and we typically choose 88% of the tachycardia cycle length at eight pulses at a time, and we'll kind of make adjustments based on, you know, how we're feeling maybe that day. So, here's an example of ATP, also hot off the presses. This is a patient of mine who also has an apical aneurysm after an apical infarct, and has this slow ventricular tachycardia. So, she goes into it, sometimes she self-terminates, so we let her go for sort of a longer detection, but this time she didn't terminate on her own. This ATP is delivered, in this case we chose 12 pulses, and then you can see that after the last pulse is delivered, she terminates. And she feels the ventricular tachycardia sensation in her chest, but she doesn't get the big, you know, whop. It's a painless therapy. And that's the ATP being delivered to kind of enter that excitable gap and extinguish the short circuit, and there's successful termination. So, after ATP, if that's not successful, then we rely on good old-fashioned, you know, high-energy shock. So, what happens there is that there's high energy delivered to all the myocytes, they all kind of get, you know, sort of depolarized all at once, and the idea is that all the cells will then reset simultaneously, and ideally sinus rhythm will take over. So, you know, how do we, you know, program these? We choose a level of energy, so we can choose a relatively high level of energy or a lower level of energy. It turns out anything over five joules, patients generally perceive the same way as being kicked in the chest by a mule is a frequent descriptor. But if the rhythm is organized, oftentimes we can deliver a lower energy shock, which will not extract as much energy from the battery as a high-energy shock, and our thought is that a max-energy shock will take about a month off the battery life. So, some of these patients with VT storms will come in and they've, you know, burned through their batteries. We can choose polarity, meaning that do we shock first from the coil to the can or can to the coil? We can sometimes program the polarity to, you know, change after the second or third shock, because if it hasn't worked for the first two, let's try something different. And in some devices, you can actually reprogram this biphasic waveform, meaning that you can change the duration and the tilt of the shock, which sometimes can help you get around high defibrillation thresholds. So, here's an example of a shock. This is actually the same patient who I showed earlier who had that non-sustained ventricular tachycardia. In this case, unfortunately, he developed sustained ventricular tachycardia. He gets ATP here and here. It's not successful, and so he finally receives a 36-joule shock, which fortunately resulted in termination of his ventricular tachycardia. So, what are the challenges with defibrillator therapy? Again, I think we've already discussed this. Number one is inappropriate shock, which is ultimately, you know, signal analysis, okay? It's signal analysis by the device. And, you know, what I tell my patients is that, you know, the devices are pretty smart, and thank God we have them, but there are other ways in which they just, I mean, we can look at an event and immediately know that a shock is not required, but we're not there in terms of the defibrillator analysis. And the problems we have are atrial fibrillation with rapid ventricular rates, over-sensing of other cardiac signals, including T-waves, sometimes P-waves, or pacing stimuli. Electromagnetic interference can be a problem, so we ask people not to use their electric razors, you know, over their chest. Or a lead fracture is another problem. And as we discussed, there's just an overwhelming excess of data coupled with this dearth of qualified personnel to help us analyze everything and organize everything. So, I'm gonna show you some examples of inappropriate shocks to kind of demonstrate to you that, you know, these are signals that are just, they are so obvious to us, yet somehow not yet obvious to the defibrillators. And maybe you will see these and something will click and you will figure out a way to reprogram those defibrillators to not deliver these shocks. So, this is a patient of mine who unfortunately received a defibrillator for hypertrophic cardiomyopathy, unfortunately had a lead perforation and developed a pericardial effusion. We drained the effusion, you know, the hole closed up, the effusion resolved, we took it out and, you know, thank goodness. What a trial for him. So then he goes home and, you know, calls me about seven hours later because he's just gotten a shock. And the problem is that the pericardial effusion put him into atrial fibrillation. He developed, you know, these adrenergically driven, incredibly rapid rates higher than our 200 beat per minute cutoff. And when we look at this, what we can see is here's sinus rhythm and the QRS has this appearance. And when he goes into his AFib, there's some variation, but it's clearly, you know, AFib with rapid ventricular rates that are driving the high rates. But the device can't figure that out and delivers ATP with futility. And then finally, he ultimately receives this shock. Another example is T-wave over-sensing. Now this is a plot of the cycle lengths. And just by looking at this alone, I can tell you it's T-wave over-sensing because there's a fixed variation in the R to T and then the T to R interval that's, you know, varying every other. And this is sort of this classic train track appearance. And, you know, these are pixels. These are just pixels, you know, on a page and feel supremely analyzable. And if you look at the intracardiacs, you can see indeed that he's just in sinus tachycardia. His T-waves have now come over the threshold of detection and the device is inappropriately sensing, you know, those T-waves as ventricular events. And unfortunately, he ultimately receives a shock. What about electromagnetic interference? Again, this is a very unusual case of a woman who has a left ventricular assist device was actually having gynecologic surgery. So our standard is that if the procedure is happening below the belly button, that that is far enough away from the device that nothing has to be done with the device in terms of turning it off. If you're having, you know, radio frequency applied to your left neck, then the device should have a magnet over it to kind of suspend the therapies. So this woman was having gynecologic surgery and it turned out that she had this enormous aortic stent and then she had a left ventricular assist device. And between the two of those, sort of created an antenna that allowed the signal from the cautery, you know, in her cervix to be picked up by her defibrillator. And all this electromagnetic interference was perceived by, you know, the device as a ventricular tachyarrhythmia. Whereas, you know, again, like within less than a second by looking at the shock EGM, you can see that the patient's in sinus rhythm. But the device, again, did not figure that out. And at this point delivered ATP, which then unfortunately put her into ventricular tachycardia at which point, you know, a few seconds later, the gynecologist was a little surprised when her defibrillator went off. So you don't see what happens, but the cautery stops after that. So actually it stopped there. They must have figured out she went into VT. So, you know, the bottom line is that these signals to us feel very, you know, instantaneously analyzable. And if there's a way to somehow build that into these ICDs, we'd be forever grateful. Here's an example of lead fracture. Leads are basically comprised of these slinky-like, you know, multifiler cables that go down front to the end of the leads back up to the device. And you can see how, you know, these pieces can fracture. And what happens is as those metal pieces, you know, come apart and come back together, they create this chatter that again is misperceived by the device as a ventricular tachyarrhythmias. And in this case, the patient again received an inappropriate shock. So challenges with defibrillator therapy, the inappropriate shocks, the deluge of data, and the shortage of qualified personnel are the biggest problems I think facing us. So what are the solutions? Where do the solutions lie? And I think there are other areas of medicine that can, you know, help us think about this. We discussed it multiple times, but again, this is ambulatory cardiac monitoring where deep learning was applied to millions of segments of rhythm strips. And this algorithm outperformed 30 board-certified cardiologists in terms of accuracy of rhythm analysis. Now, okay, I won't, okay, I'll leave it there. So, you know, another example I think we're all familiar with is, again, pixel analysis of CT scans. And these deep learning algorithms have been shown to be more specific, more sensitive than board-certified radiologists in terms of detecting, you know, tumors, you know, as true, you know, malignant masses versus something benign or not even recognized. And, you know, I think that, you know, EP and specifically devices, defibrillators are ripe for this kind of analysis in terms of decreasing inappropriate shocks, decreasing the demand for human analysis, more accurately diagnosing arrhythmia versus lead problems or, and maybe even improving detection or prediction of device malfunction. You know, this is a lead that has impending lead failure as opposed to, you know, discovering it after the fact and helping us anticipate, as Dr. Russo was saying, the development of arrhythmia. Thank you.
Video Summary
The video focuses on defibrillator therapy, specifically the two types of therapies that can be applied: fast pacing or anti-tachycardia pacing, and a shock to the chest. The video explains how defibrillators work by delivering a high-energy therapy to the heart muscle through electrodes. It also discusses the programming options for the devices, including the cycle length, duration, and waveform matching parameters. The video highlights the challenges of defibrillator therapy, such as inappropriate shocks, over-sensing of other cardiac signals, and lead fractures. The video suggests that solutions to these challenges may lie in the field of deep learning, similar to how deep learning algorithms have been successfully used in other areas of medicine for accurate diagnosis and analysis. The video concludes by suggesting that applying deep learning algorithms to defibrillator therapy could help decrease inappropriate shocks and improve the diagnosis and prediction of arrhythmias and device malfunction.
Asset Caption
Susan S. Kim, MD, Northwestern University, Chicago, IL
Keywords
defibrillator therapy
fast pacing
anti-tachycardia pacing
shock to the chest
deep learning
Heart Rhythm Society
1325 G Street NW, Suite 500
Washington, DC 20005
P: 202-464-3400 F: 202-464-3401
E: questions@heartrhythm365.org
© Heart Rhythm Society
Privacy Policy
|
Cookie Declaration
|
Linking Policy
|
Patient Education Disclaimer
|
State Nonprofit Disclosures
|
FAQ
×
Please select your language
1
English