Josh Lupton, MD: Amiodarone timing in Arrest

I don't know what. That works. It's recording now. Go ahead.

Yes, yes, yes, yes. Awesome. All right, Andrew, thank you so much. All right, Everybody, welcome. It's 3:03pm on another Tuesday here for the Florida EMS webinar, where we've been very grateful to have some incredible speakers from around the country and really around the world join us to share with us the information that they've published. And today is no different. We have Josh Lupton from the Oregon Health and Science University, OHSU, and he works there with Dr. Daya, who has been on this webinar many times, not just as a speaker, but as a participant. And I ran across this paper.

Recently, maybe a few months ago, and the paper is entitled Survival by Time to Administration of Amiodarone Lidocaine or placebo. Placebo in shock, refractory, out of hospital, cardiac arrest. So obviously something that is very pertinent to what we do in the field. And I thought it'd be a good kind of look behind the curtain of what the data shows on this topic. So without any further ado, Josh, thank you for joining us and please take it away.

All right, perfect. Thanks very much. And feel free. With teams, it's always hard to see a chat or anything while I present. So feel free to interrupt if there's any questions or we can save them to the end whenever everybody prefers.

So this was a paper we published, was available back in March in Academic Emergency Medicine. I'm Josh Lupton over here at ohsu. This research specifically was funded as part of a research training grant. So this is one of the aims of the grant from SAEM Foundation. Otherwise, no other disclosures. So, as many of you know, for a shock refractory arrest, particularly one with an initial shockable rhythm, that's what you see on your initial rhythm check. You give your first shock, there's no drugs, your second shock, you then give a dose of epinephrine and you continue on. We locally give amiodarone. ACLs would say amiodarone or lidocaine after that third shock. Then again, if you're persistently in a refractory arrest with a shockable rhythm, you'll give another dose of epinephrine, your second dose of amiodor and lidocaine. This is pretty consistent across the European guidelines as well as the AHA guidelines, though the European is a little bit more specific to amiodarone.

Then when we think about, well, what in practice, how long does it take to do these different interventions? There's been a nice aggregate study of these times from Joe renato Back in 2021, that looked from the 911 call and what we see is that a shock occurs at about 8, 11 minutes. These are all median times in different studies. Then the first dose of epinephrine is about 15 or 16 minutes, whereas for amiodarone it's about 20, 26 minutes. And the 26 minutes is in more of a registry based, whereas the other three are in these trials where they're specifically giving antiarrhythmics. And so one would imagine a little bit more inclined to give it quicker relative to the arrest period. And so in noticing how long in our local data and then in this published data it takes to sort of get these antiarrhythmics on board, our question was, well, what if you give it a little bit faster, could that sort of improve outcomes? This sort of goes in line with this three phase model of cardiac arrest that many of you know. And in this instance, an example of someone with an EMS or, excuse me, a bystander witness arrest, you get this 911 call. Well, it's early electrical phase. I always like to say. It's really the time where even if you didn't do cpr, if you shock someone in those first few minutes, they're going to ultimately have a good outcome. They're going to return to a perfusing rhythm, most likely, hopefully right. For someone with shock refractory arrest, that electrical phase really continues throughout. You ultimately get to a circulatory phase where you need cpr, maybe some epinephrine and then eventually metabolic phase. Well, when you overlay the time to our Drugs from that 911 call, you see that the majority of these antiarrhythmic drugs which really are attacking that electrical phase are given fairly late in this process.

As we think of trials and interventions in general for cardiac arrest, this schematic sort of has helped me to get the idea from arrest to, to ultimately death. There's sort of this window here and you have patients who again are likely to have a good outcome. No matter what you do, the arrest is too short, that if they get rosc, whatever your intervention is, those in the control group and the intervention group are both going to look pretty good. You then have the patients on the far right, which those that have been in arrest for so long, no matter what you do, there's really a point of futility in that group. And even if your intervention might be the best in the world, it's not going to have a good outcome. And I like to say if this was to test, say, defibrillation for shockable rhythm, and you waited for an after patients had CPR before they were randomized to be defibrillated or not, well, majority of those patients are not going to do well. That's kind of an extreme example. But what we wanted to know for antiarrhythmics was is there more of a critical window? And when you think of sampling for a trial, if there's too many in this sort of futile group, that gets drawn in. So those that get the randomized intervention too late in their course of arrest, are you going to somewhat bias the results to the null? You see here, this blue group has a 20% absolute benefit. But if there's so many in the futile group, it can really draw that down compared to if you look at patients who are more from this critical window group.

That's what really drew us to want to do a secondary analysis of the amiodarone, lidocaine and placebo trial. Many of you, of course, are very familiar with this trial. It was a randomized, individually randomized trial, not cluster randomized like many of our more recent studies. It was double blind trial. And it compared amiodor, lidocaine or saline placebo for shock refractory arrests. In this trial, patients had to be refractory to just one shock. So patients in theory could get the drug after two shocks, which is a little earlier than our current ACLS algorithm, and they could receive it via an intravenous or intraosseous access. And our population that we looked at was the same as the primary population from the parent study, which was those with an initial shock Alarithm who received the randomized drug. You see the ultimate primary outcomes overall, which was survival to discharge and was not significant between amiodarone or lidocaine compared to placebo. And I believe the P value for amiodarone was say 0.08 in the overall group. But when you look at the preplanned supplemental analyses from this parent publication, you'll see that in EMS witness arrest cases, as many of you know, the absolute benefit was about 21% or a number needed to treat a little bit under 5, which was fairly significant for EMS witness. For bystander witness, it was still significantly higher survival discharge for those with amiodorone versus placebo, but much less so. And then those with unwitness, which I think you would agree are the ones more likely to be sort of in that far right of the schematic in the futile category, had no Significant difference across the, across the board. And so our question with this was, well, perhaps regardless of witness status, those that get the drugs earlier in their course of care may have a better, more time for the drug to work before that futile period. So our objective was to look if early anti arrhythmic administration was associated with a higher survival and really a favorable neurologic outcome survival compared to that placebo or normal saline arm analysis. What do we use to define our early drugs? Before we had access to the data, we decided that we wanted to use time from ALS arrival on scene. The reason for this over the 911 call was this. The idea that something needs to be done closer to the 911 call is sort of convoluted by the need to get on scene as quick as possible, which we know ultimately improves outcomes. And though you certainly could control for that in that analysis, we really wanted to have more of a hard number of from when ALS gets on scene, you know, how does that impact the overall outcome? So we decided eight minutes was going to be our cutoff. We figured that allowed a couple of minutes for that first shock and then around those two minute intervals and then another two minute grace period after the third shock to get antiarrhythmics on board. Again, this group would encompass patients that could get it per the current ACLS guidelines. After the third shock. Ultimately only 1/3 of patients met this early category, with the remaining 2/3 getting the study drugs after that 8 minute cutoff. That was our time based definition. And then we again planned to do more of a sequence of care based definition for this. We wanted to look at those that got the study drug near the initial dose of epinephrine, used a one minute caliper. And then we essentially said that those patients got it early if they got it within one minute of that first dose of epinephrine, otherwise they received it late. In our late category.

We used the same per protocol population so that 3026 we used multiple imputation for some missing variables to help reduce some of our bias. Though all of our results were the same in our complete case analysis. And then we specifically excluded patients who achieved ROSC before the study drug. And the goal of this was that if you had single shock, you would rosc for say 20 minutes and then you rearrested and got the study drug at that time. You know, your time to drug is going to be very long, but your physiology is a lot different than those in a true refractory arrest, at least up to the point of getting the study drug.

We controlled for all of the typical variables in our analysis. And I don't have the table one here, but you know, in terms of things like sex and time for EMS to get on scene and witness status and bystander interventions, they were fairly similar, similar across the groups. And though this is a secondary analysis, because it was an individually randomized trial and it was double blind at the individual level, you would expect when you do a secondary analysis across those, you know, the hope is if the parent trial was adequately randomized, you should have somewhat still minimal differences between your groups. So in our unadjusted and then I'll show our adjusted results, what we found is that whether using a time based definition or sequence of care based definition, those who got early study drug compared to those who got late study drug, there were no differences across what you were randomized to. In the late administration groups and in the early administration groups there was a consistent difference whether it's survival, hospital admission, survival, hospital discharge, or many of us care about a lot, which is functional survival at discharge, where amiodarone compared to early placebo had significantly greater. Again, unadjusted survival to discharge with a good neurologic status, There were no differences in the lidocaine compared to placebo group. In the late group, there was no really differences across all three drugs that patients might get.

This was our unadjusted. We wanted to adjust for all those confounders mentioned, which we additionally included time to initial defibrillation attempts, time to access and time to first dose of epinephrine with the idea that perhaps those that got antiarrhythmics earlier, maybe they had faster defibrillation, faster this or that. Now you would think that those benefits would be equally felt across the randomized arms, but we wanted to be sure. What we saw is that highlighted here in the lighter colored bar which is on the left, again the time based early cutoff on the right, the with epinephrine group, there were significantly higher adjusted odds for survival to hospital admission, survival to hospital discharge, and again functional survival in those that received amiodarone compared to placebo if it was given in this earlier window, this critical window, and that's the same as if it was given with epinephrine. And not surprisingly and again indicative that the parent trial had good randomization. There's not a huge difference in our unadjusted and our adjusted effect size. We really don't see any differences in the late group, particularly for survival to hospital discharge and functional survival to hospital discharge. Again, similar to our unadjusted analysis, then our study is generally.

Underpowered to detect significant interaction, which would require a significantly higher number of enrollments. But even with that, we see this P value for interaction by time of of drug delivery. So the early versus late category is significant in some instances with survival to admission and near significant for the other, which again just suggests that there might be some differences in efficacy by time of delivery.

This here is our continuous results. And so you see on the left, time from ALS to drug administration. And then as part of a sensitivity analysis, we looked at time from arrest. So if the arrest was witnessed by a bystander, we use the 911 call, and witnessed by EMS, we use the true arrest time. And what we see in both groups is there's not a clear cutoff in terms of where there's. In terms of. On the left side of the graph.

It seems that the earlier you go to administration, the greater the difference between the amiodarone versus placebo arms. But when you look to the right, you see that at some point, all of these lines really intersect. And that sort of fits with our kind of theory that maybe after a certain period of time, there really is just a period of futility. In this trial, at least in the comparison between drugs and placebo, I won't spend much time on this, but we tried other graphs for modeling besides just linear to take to account that maybe the slope of change of this effect would be different in the first few minutes between minute 5 and 10 and after minute 15 or 20. But none of these were significantly better than linear modeling, so we stuck with that.

Then we did the more classic analysis where rather than compare to the placebo arm, we compared each drug as given, giving it one minute later. What you see here is that even with adjustment, giving placebo one minute later has a 10% decrease in odds of functional survival. This is really indicative of the resuscitation time bias, where in many studies where you don't have a placebo arm to take advantage of, it can be difficult to account for the fact that those that get an intervention later are almost always in arrest for a longer period of time. Still, we see the magnitude of the decrease per minute is much greater for amiodarone than placebo, which is likely indicative of the effects I've shown you previously.

How does this compare to other studies that have looked at similar things? You see this one at the top in 2023, which was a study from an Australian retrospect cohort, they did time based matching since they didn't have a placebo arm to account for this resuscitation time bias. And they looked at all different quartiles of time to matching or time to drug delivery. I'm just showing their 25th percentile. So 25% of patients received it within 18 minutes of the call and they saw a significant improvement in the or significantly greater.

Survival to admission and survival to discharge in terms of risk ratio in those who received antiarrhythmics early compared to those that did not receive them at that time. This analysis excluded times where patients had less than four shocks, which I think is very fair thing to do, but may undercount the benefit. If someone gets an antiarrhythmic after two shocks and doesn't go on to need four shocks and has ultimately a better outcome, in theory, they would then be excluded from sort of this analysis. And then this other study that had looked at this in terms of neurologic outcome found that those who got antiarrhythmics within 20 minutes from the call had about a 10 greater odds of a favorable outcome. But again, this is a study that didn't account for that resuscitation time bias. So likely is, let's just say, an extreme version of the potential effect size.

So I think taken together our conclusions are really that earlier amiodarum seems to clearly be associated with a significantly greater odds of a favorable neurologic outcome than early placebo and shock refractory arrest. This wasn't the case for lidocaine delivered early in this group. Also notable is that it ALPs only 1/3 of patients really received it within 8 minutes of when ALS arrived on scene.

Then the only other thing to highlight here in the last minute or two, so we have time for questions, is this question of route of administration, which is something Modaya myself and several other co authors looked at back in 2020, also with the ALPS data. What we did then is we looked at those who received the drugs intravenously versus intraosseously, again across these randomized double blind arms. What we found is that those that received intravenous antiarrhythmics, both amyo and lidocaine, had significantly greater outcomes compared to placebo than those who received it intraosseously. One potential theory is that particularly amiodarone but also lidocaine, are very lipophilic. In the bone marrow space there is a high degree of lipids and there's been some studies to suggest that you can get some of those lipids even in the serum Circulation. So there may be some sequestration of the F efficacy of these drugs through the intraosseous route. But In Alps, about 97% were tibial humeral. Intraosseous routes generally have a lower fat content, and so it's tough to say if this effect would be the same. We looked at it though, with that in mind, and what we found is that intravenous really mimicked our primary result, which since 75% of the route in Alps was intravenous, that's not surprising. But when we looked at the intraosseous group, we were wondering, well, what if you give it early enough, the benefit of an IO maybe is that you can place it faster. Do you see sort of a more of an effect in that earlier group? And I would say we really did not. So this is kind of similar to the other publication where in the IO group there was really nothing significant. Even despite the low numbers, the effect sizes were not particularly suggestive of a true benefit.

And so what does this mean? What are the future directions? Well, I think, you know, no one would argue if you have an IV in place, that's probably the best route, but it may not be as easy to place quickly. Then I think from the data we've presented here, from the other studies, and from the lack of published data that early antiarrhythmics are harmful, particularly in cardiac arrest, perhaps earlier is more optimal. Then the question is, well, how do you move from later administration? Is an IO overcome? I think our results here would suggest at least tibial does not. But would a humoral site balance that out? Would a higher dose of these drugs balance that out? Really, ultimately, the goal is to try to get patients to resemble that EMS witness group. They got antiarrhythmics fairly short after their arrest. And compared to those that got placebo in that EMS witness group, those that got amiodarone, the number needed to treat was about 4.6 for an extra survival to discharge. We are trying to look at this locally, prospectively, which is another part of this grant. This is just to give you a sense of our times from arrival on scene interventions, which are very similar to those in Alps, which is that the times between shocks often stretches a little bit longer than 2 minutes. And our median time to antirhythmics about 11min. Again, about 1 third of our patients get it within that eight minute period. But in our data set now we're using about a little over half. In 2022 had humoral iOS, much lower proportion of tibial iOS overall and a lower proportion of IV than in ALPs. But ALPs had very little humoral IO. So it'll be interesting to see if these sort of trends hold. When we started to look at this and presented this at SAEM here back in May of what we found is that in patients who all required at least three shocks, so trying to account a little bit for that resuscitation time bias, those that received amiodarone between shock 2 and 3 compared to those that were eligible but didn't receive it between shock 2 and 3 and received it later, the intravenous arm had a significantly higher adjusted odds of survival of discharge, sort of similar to that other publication that had an odds ratio of 10. But the intraosseous did not have that same effect and really resembled more like the placebo results from our ALPS analysis. And there was a near significant interaction between routes and antiarrhythmic timing. This is in our retrospective group that were back in 2018, 2020, still low humoral use. And so we're looking at this prospectively, trying to also take into account if humoral is a potential avenue to alleviate some of the.

The less of an association with benefit that we see with the tibial iOS. That's the whirlwind of the optimal timing and rep studies we've done with antiarrhythmics. I think we've got a few more minutes. I will stop sharing so I can take some questions.

Josh, that was incredible. I remember when Mo presented at the podium that paper on IV amio versus IO and so forth, and I remember leaving that presentation, like, upset, like, how did he come up with that? You're.

Oh, I think I lost him. Can you all hear me?

Yeah, we can hear you. He.

Yeah, we lost Peter as well. Okay, Josh, you're still there. I think I'm still here. Yeah, you're still live. Perfect.

So, Josh, just real quickly. So in a sense, what we're hearing is that the sooner the better. Right? Particularly for the amiodaron, but not so for the lidocaine within the numbers that you were able to look at, right?

Correct.

Yeah. Within the confines of alps, and they had the same number of lidocaine and amiodarone. So you would anticipate seeing a similar effect size be significant. There was. There was not the case for lidocaine at either of the time points. But for amio it seemed sooner the better or at least within that eight minute cutoff.

And that's what drove the whole lab study to be positive, especially in the witnessed arrest, which makes a lot of sense and all that.

So.

Okay, well, thank you so much for the great work and all the analysis. It's painful going through that, but. But when you're out there actually seeing the patients, you can see this happening in real life. So it's really great to have the numbers corroborate that. So.

Yeah.

Hey, sorry, guys, I. My computer just completely crashed on me. Sorry about that, Paul. Thank you. Thanks for filling in there. I didn't hear your question, Paul, but.

It wasn't very. It wasn't very good anyway, so just.

No, but what I was going to ask before I got swiped away there, Josh, was the theory you have about the high lipophilic content of the marrow and it being less so in the humoral head and so forth. Has that been evaluated anywhere else or. Because this is the first time I'm hearing that. I think it's a fantastic theory. Has that been looked at in any other, like, pharmacologically?

Not particularly. I think most of the pharmacologic studies, so they've done some in swine models with antiarrhythm, usually with tibial, sometimes compared to sternal, which is not quite as comparable to humeral, but fairly similar. And it looks like there's a little bit of a delay in some of the parameters, meaning, like the time to max concentration and what the max concentration is for tibial, comparatively.

And so there's some suggestion of that. But there's been no human studies of sort of the pharmacokinetics for something so lipophilic like amiodarone. They've done it for a lot of other substances that are very similar in efficacy. And they, I think, most recently the calcium trial, you know, looked at the calcium concentrations for those that received it intravenously versus intraosseously and found that they were comparable in the intervention arm and comparably lower in the placebo arm for both. Right. And so, as one would imagine, things like epinephrine, calcium, those sort of. Of drugs, you could convince yourself that in a lipophilic environment they might be equally efficacious. And so an IO might be appropriate for those. And I think similarly, and.

Josh, can I interrupt you because I have to go here pretty soon.

Yes.

The other issue is the preparations that were used, whether it's an aqueous amiodarone or the lipid. Philip 1. So just let them know about that as well, because that did affect the studies as well.

Okay.

Yeah, for sure. Yeah. So they use this captisol version in Alps, and. Which is not the one typically used in the bolus cardiac arrest dosing currently or since. And so the. How different the lipophilic properties of that comparatively? I mean, amioderm by itself is fairly lipophilic. So it's not that it's all sort of the formulation. There's some other side effects of the formulation, differences.

That make it harder to take ALPS data and apply it to the formulations of amiodarone we're using regularly. But that's somewhat of a whole separate can of worms to open.

Exactly. But just want people to be aware of that, that's all.

For sure.

Yeah.

Yeah.

Interesting. So the other question I have for you, we have removed EPI from our shockable protocol. Shockable rhythm protocol. I wonder. Let me just. My computer's still acting up over here. You know, removing the epi, therefore pushing up the meo. You know, I wonder, I wonder if, you know, we ought to be looking at our numbers now based on that. That's interesting, but have you guided your paramedics to modify anything specifically.

Either IV or IO in your agency when they give the amiodone based on your. Fine.

I mean, our goal is to give it intravenously if it's available. There's not a specific change. There's not a specific delay protocol, meaning if you only have an IO, you know, you still would give it intraosseously if it's. If it arrives at that time. But our goal is to try to do some studies here where we change up perhaps that algorithm, if that makes sense. And I think the way I've explained it to folks, too, you know, if you have stable vtac, you know, the last thing you're going to do is give that patient a milligram of epinephrine, and then. And then if that doesn't work, you know, give them an antiarrhythmic. You know, usually you're. You're reaching for your antiarrhythmic sooner in your stable vtach patients. So when they become pulseless, it's strange, just sort of drop it down a little bit. And that sort of kind of drives the, the thought process of our, of our work that the different formulations of amiodarone, you know, the, the one that's currently available theoretically, could cause some hypotensive effects. And so, so I don't know if we would drop off epinephrine from our protocols if we move up amyo or just give it with epinephrine, those sorts of subtleties. There's not good evidence for One or the other, right?

Yeah, of course. Anybody else with. This is fascinating, let me tell you. I mean, it really. You got. Someone raised their hand. This would be. Go ahead. And whoever's raising their hand, I can't see on my phone, but go ahead and ask your question.

Hey, it's Lou Steinberg. Just a quick question and you'd have to think about it. With distal humeral becoming more popular, would we think that would be more comparable?

What do you mean, Lou? I'm sorry, did you mean distal femur?

I'm sorry, a distal femur, would we expect that to be more comparable to humeral or tibial? As far as the administration, yeah.

I mean, I think there's a few different reasons why lower extremity might be worse. So not just the lipophilic. The thought is that the IVCs don't quite have the same valves. And so in a low flow state of cardiac arrest, the question is, is it just the physical location that there's less return of that blood flow without those valves and so your drugs get less circulation? And then the other would be. Then this. This lipid sort of sink, you know, theory of lipophilic drugs. Femur, I would imagine. I don't know offhand, but I don't view that offhand as a lower lipid content, sort of long bone. So I don't have a reason to think it would be better or worse. But quite frankly, like, I'm a big believer in, like, you know, the trial. The trials need to sort of be done to say what the answer is on both access. And then once we kind of sort that out, perhaps timing. And I think the end result may be that the route of access matters based on the rhythm and the drugs, whereas.

It may just be more important to get a route that you can get EPI in right away.

But yeah, it's funny, actually, the folks in Salt Lake, they actually did a 5 milligram IM. You saw that. Scott Young. I mean, in their small sample, they found that there was no difference. So. Fascinating stuff. Wow, this is. This is just great to hear it from you because, like, your explanation really brought it to life. Everyone got that same feeling. Anybody else with any last question before we hop over?

All right, so, Josh, give us the last word.

What are your.

What would you tell us here in Florida? We have people joining us from around the country as far as things that we should be considering, you know, just as far. I know you kind of gave us the overall view, but what would you leave us with. With respect to amiodarone and cardiac arrest.

Yeah, I mean, I think my parting wisdom from looking at it is that sometimes the. The location of things in the algorithm for cardiac arrest are more because that's where they've been and not because of evidence of harm, of slight modifications for those. Both true for EPI and antiarrhythmics. And so, again, I think for amiodarone, it seems like earlier might be more efficacious without good evidence of harm. And so I think I'll leave it at that for what folks want to do, but it probably needs to be studied in a trial, which we'll try to do on our end. If we can get funding for.

Well, let me tell you, you know, this is the kind of evidence that really is going to move our forward and get more people.

And get more people back to life. So, Josh, thank you for joining us, and thanks for your contribution. Thank you so much.

Absolutely. Thank you all. Take care.

Get it. All right, Joseph, you've been so kind as to take us on over to Dr. Pepe. I appreciate that, and I'll see you guys on that. Thanks, Joseph.