I appreciate you attending, I'm Casey Bryant, Dr. Gordon, Dr. Ishida, so Dr. Gordon is from Temple and is a pulmonary intensivist that has some specialized practice in post-cardiac surgery care. Dr. Ishida is from NYU Langdon, is a stroke specialist. We're going to talk about some issues that you run into after cardiac surgery and my portion is going to focus on vasoplegia and RV failure. It's a 20 minute session for me so obviously I'm not going to be able to cover everything but we'll try to hit some high points and send you home with some pearls and hopefully you'll learn a little something today that will help your practice. Disclosures I have given talks for Fisher & Paykel before, it's not relevant to this talk and I'm coming from Wake Forest in North Carolina. We do have some audience response questions in there so if you want to scan that QR code and we're going to start with one as well so if you don't get it now you can get it in just a second. So the objectives for my session will be covering some high points on RV failure and post-operative vasoplegia. As I said I can't cover everything, we're going to leave out onotropes because there's no new data on onotropes and I'm not going to cover mechanical RV support either because that data is terrible but it is something that we do. So here's the first case, we've got a 68-year-old female who has a combo CABG valve surgery, two hour bypass time, 80 minute cross clamp time, standard medical conditions, has some mild to moderate RV dysfunction coming off a bypass, is placed on some dibutamine, the pain is adequately treated post-operative and you notice that the PA pressures are a little bit elevated and you want to start an inhaled pulmonary vasodilator so my question is what is your choice? One vote, there we go. So we've got actually a majority are going with inhaled nitric oxide and then you have the other choice for inhaled ibuprofenol, inhaled nitroglycerin, that's something that you can do, it is an inhaled pulmonary vasodilator, there's not a lot of literature in the cardiac surgery world for that, PS sildenafil obviously you're going to get some hypotension systemically from that so that's not desirable post-operatively. And then there's the option to go combo therapy which we've done at Wake Forest, the data is terrible and it's usually just clinician preference, there's no data actually. So I would say you can go with either A or C and we'll go into a new study that recently came out this year that may help guide your decisions moving forward. So when you talk about RV failure, I think you know it when you see it but you're talking about a low cardiac index, low mixed venous and then the need for RV support. And we all know that the RV especially if it hasn't adapted to pulmonary hypertension over time has a difficult time adapting to any acute change in the afterload and if you experience RV failure after say an LVAD implantation that takes your mortality risk almost up to a third. So it's a big problem and what we all want to avoid is that RV spiral of death where you have RV ischemia, RV wall dilation leads to decreased output, more hypotension leads to more wall distention, more ischemia and you get into this spiral of RV death. The burden, it's about a third of cardiac surgery cases, there's variable definitions but depending on the definition you look at we see it about a third of the time. The causes, I'm not going to read those to you, I'm just going to point out they're all due to the surgeon. I am an intensivist, I'm here to blame the surgeons for everything. Do you have any surgeons in house by the way? Any surgeons here? Perfect. All right, it's a safe space too, that's good. So we have echocardiographic indicators, your TAP-C, S-prime, RV fractional area of change. There was a talk earlier this morning that went into all of these, it was very good so if you want to go into detail on those things go back and review that talk once it's, you know, the recording is published. You have your hemodynamic parameters as well, if you have a rising CVP, decreasing cardiac index, your CVP is rating your wedge and then we have the PAPI score. So the PAPI score actually has some new data that has come out since 2020 and they were using this to see if they could use it to both predict and identify RV failure in the cardiac surgery populations. And they found that if you had a PAPI preoperatively around two, that that was predictive that you would have postoperative RV dysfunction. It's a very easy score to calculate, you have your pulmonary artery systolic pressure minus your pulmonary artery diastolic pressure divided by your CVP. You can see this has a really good area on the curve of the ROC, so .8, pretty good sensitivity, very good specificity and they found their optimal cutoff was about 1.9. So if you have a cutoff of about 1.9, then you're going to very likely deal with RV failure and dysfunction postoperatively. The lower it is, obviously the worse it is and I feel like this is a very useful tool because it's easy to calculate and, you know, you can do it, you know, throughout your time caring for the patient. So if you use it, you know, before the operation, you can predict it. If you use it after the operation, it will help you identify that RV failure. And then when you make interventions, it can help you assess whether those interventions are making anything better, which I think, you know, should be your general approach when you add any type of intervention to your patient is, I had a problem, I had a, I developed a hypothesis, I implemented a treatment, circle back, did that treatment make anything better and do I need to change something? So in general for RV dysfunction postoperatively, we have a kind of three pronged or three legs of the stool approach to supporting the RV. The first is reducing the pulmonary vascular resistance and that's, you know, multifaceted, so it's avoiding hypoxia and hypercapnia. It's trying to establish, you know, your FRC on the ventilator but not over peeping the patient. So this is, you know, various aspects of care that go into reducing pulmonary vascular resistance but you can also use inhaled pulmonary vasodilators. You want to improve coronary perfusion, obviously the RV is perfused during diastole primarily and that may be just you treat hypotension but it could mean you go as far as putting in a balloon pump to augment that diastolic perfusion. And then the last component is optimizing preload, which is, you know, something we're always chasing but in the postoperative period when these patients are, you know, in the vasodilatory and capillary leak stage of their recovery from surgery, it usually means giving them volume in that phase. We all know that post-cardiac surgery RV failure is different than someone who has an MI and has RV failure from that or somebody that has a PE and has RV failure from that. And then improving contractility and that's generally, you know, adding on some sort of inotropic support and as I mentioned earlier, that data is pretty much non-existent for post-cardiac surgery care. So your two agents that most centers use would be inhaled epiprocinol and inhaled nitric oxide. They both do the job. They both lower your mean pulmonary artery pressure. They both improve VQ matching. They're safe. They wrap it on, wrap it off. They avoid systemic hypotension. They have two distinct pathways and this leads to a lot of clinician preferences. Well, the nitric oxide is better. Certainly their marketing is better. That Knox box looks really cool. But there's a lot of clinician preference that goes into which one of these agents that you select. So this study was actually published by one of my colleagues at Wake Forest, a single center retrospective and, you know, looked at primary outcome, did it lower mean pulmonary artery pressure? Yes. But then all these secondary outcomes, there was no difference in any of those other than the cost. You know, inhaled epiprocinol was less than $400 to treat a single patient whereas inhaled nitric was close to $2,500. There was this meta-analysis that came out where they basically put together a bunch of retrospective data and said there's a lot of bias in the study. The population is very heterogeneous. It's hard to make heads or tails out of anything. There were different weaning protocols and that type of stuff. But on the whole, they said there was no difference between the two as far as their clinical effectiveness based on the data that was out there. Well, now we have Dr. Ghedimi and the Duke folks to thank for this recent study that came out this year in circulation. And this was an RCT where they randomized patients to get either inhaled EPO or inhaled nitric oxide. You can see they had 100 patients in either arm. And they were looking for composite progression to RV failure. And that RV failure was defined as if you required an RVAD, if you had an open heart transplant, or if you met Intermax moderate to severe RV failure criteria for LVADs. And they looked at a bunch of those important secondary outcomes. And they take home there was no real difference between the two agents. Neither one of those agents is magical or is going to just make your patient miraculously get better. They both did the job and lowered mean pulmonary artery pressures. And based on this study, I would say the difference between the two agents really is just called at this point. Until we have better data, I would say that inhaled epiprocinol would be a very reasonable choice. There was a signal in the inhaled epiprocinol group for more deaths. It did not reach statistical significance, but it was enough for the authors to go back and look at all these. And then obviously it was a low number of patients. They looked at them all and said, do I think that any of these were attributable to the agent? And they came away with the conclusion that it was not. So now we'll go to our second question. The core of this is basically you have a patient with RV dysfunction that now has some hypotension that wasn't responsive to fluids. So what is going to be your approach? Sorry about that, let's try to get to 10 again. All right, so we have a preference for norepinephrine. I would say probably if someone has RV dysfunction that vasopressin should be your agent of choice. And some of that is due to where vasopressin is active. And this used to just be animal data, but now we have human data as well. This was a study looking at cardiac surgery patients and what they did was they loaded them with milrinone and then placed them on a milrinone infusion. And then to counteract the hypotension that came, they either started them on vasopressin or norepinephrine. So the norepinephrine group is on the left, vasopressin is on the right of your screen. And when they started the milrinone, which is those middle data points in each graph, it increased the PVR to SVR ratio, right? So there was more systemic hypotension, right? So it increased the PVR to SVR ratio. So when you try to address that hypotension with norepinephrine, that ratio stayed high, that PVR to SVR ratio stayed elevated. But when you started vasopressin, it restored, you can see how it restored this PVR to SVR ratio. And that was because the vasopressin was active systemically, but did not add afterload to the right heart. So I would say in general, if you're experiencing hypotension in the setting of some RV dysfunction, that probably your first pressure agent should be vasopressin. Inhaled EPO and nitric oxide, they seem equal in all things but cost. That PAPI score can help you identify and predict and then assess whether or not your interventions are being helpful in RV failure. A few points on inotropes. If you get into the literature, some centers put everyone on them, some centers put barely anyone on them. I think you gotta look at the patients in front of you, think about their physiology, think about the pharmacokinetics of the drugs they're using and make a selection that you think is best for your patient. And in general, I think a couple of agents is probably better than just leaning really hard on one agent. And then always fix what you can. Hypoxia, hypercapnia, preload, hemoglobin, calcium, those easy low-hanging fruits are always good to try to address. So next question. Same patient, index is fine, but the hypotension has worsened. The hemoglobin is normal, CVP is 12, and despite being on norepinephrine 20 and vasopressin 0.04, the SVR is still 650. So what would you do next? Three more, two more, one more. All right, good. All right, so most would go with methylene blue. I think that's what most people would do. But hydroxycobalamin got a good number of votes and angiotensin too got a good number of votes as well as streptosteroids. So vasoplegia syndrome, up to 44% of patients. If you have risk factors, 30 to 50% of patients. And if you have it, it's a bad problem because it increases your mortality risk from three to five-fold, which is obviously not a good thing. And it increases a lot of negative patient outcomes like time on the ventilator, ICU length of stay, renal failure, those types of things. Index should be normal, SVR is low, and you should have fixed their feeling pressures and that kind of stuff. So risk factors for vasoplegia. If you'll look at these again, I wanna just point out that these are all due to the surgeon as well, pretty much. There's two patient ones with renal failure and previous cardiac surgery, but the rest of that is due to things the surgeon did to them. Depressed EF and diabetes had a trend towards developing vasoplegic syndrome but did not reach statistical significance. And then ACE inhibitors, which had been reported previously as being associated with vasoplegia were not associated with vasoplegia in this study. So management, I think you have two carts. You have your vasopressor cart and you have your adjuncts. And when you talk about vasopressors, you have your catecholamine group in the blue and then your non-catecholamines. And I think most people, just based on the physiology of these drugs and studies that are out there, go to norepinephrine, vasopressin, angiotensin II. So we have this study here where they randomized people to either be placed on dopamine as a primary vasopressor up to 20 mics per kilo per minute or norepinephrine up to 0.2 mics per kilo per minute. And then you see the rescue drugs there. And the thing I wanna point out between the two groups is that the dopamine group got second line drugs a lot more frequently than the norepinephrine group and also got placed on dopamine at much higher rates, almost 50% of those patients. And the only safety signal that was out there was that the creatinine in the norepinephrine group was a little bit higher. And to me, when I look at this study, I'll say, this tells me that we should probably be using a bunch of different drugs that help the patient in different ways than really just pushing the gas pedal down really hard on one single drug. General approach at Wake Forest, and this is kind of what I use, is fix what you can. Obviously, your cardiac output, acidosis, hemoglobin, calcium, that low-ingredient fruit, take care of that first. And if you're still hypotensive, then we generally go with norepinephrine first. And I usually will move down and add vasopressin if their norepinephrine dose has gotten to about 0.1. So I usually add vasopressin on pretty early, personally. And then if they're still hypotensive after adding vasopressin on or going up on norepinephrine doses, there's not something easy to fix, then I'm thinking about angiotensin II or other adjuncts. There's some costs associated with these, which is why pharmacy really restricts the use of angiotensin II. But you can see, as you go down that algorithm, the cost of things goes up. There's no relation to clinical practice with that either. I just want to say that that just happened randomly. So the VANK study, this is a very important trial if you take care of a heart surgery patient. This is our only RCT looking at vasopressin, really outside of that dopamine study. And this was looking at vasopressin versus norepinephrine. And they had about 150 patients in each group, and they were looking at this composite primary outcome of death plus a bunch of negative outcomes like strokes and sort of wound infections and renal failure. And if you get into the data, the difference between the two groups was basically all because of differences in acute renal failure. And the way they defined it was creatinine greater than two and at least double their normal creatinine or needing renal replacement therapy. So the difference in the outcome, this composite outcome, was all due to renal failure, no difference in mortality. But renal failure's an important patient-centered outcome. The norepinephrine group had more AFib. That's not surprising. There's some other secondary outcomes. I don't know what to make out of. But then they measured blood levels of vasopressin, and not surprisingly, the vasopressin group had vasopressin sustained out longer into their stay, whereas the norepinephrine group initially had a rise and then had a fall. And a relative ADH or vasopressin deficiency is felt to contribute to ongoing vasoplegia. As far as the renal failure, it makes sense that you see better renal outcomes with vasopressin when you think about where it's active, being after the glomerulose versus norepinephrine that's active before. So now I'm gonna move on to angiotensin II. We're gonna try to discuss this briefly. So we all know angiotensinogen to renin, or by renin to angiotensin I, then by ACE to angiotensin II. If you get endothelial injury and you block ACE, then basically you get diverted over to this angiotensin I to VII byproducts and you start accumulating bradykinin. And all of those things are vasodilatory, which is not a good thing when you're dealing with vasoplegia. I think everyone should be familiar with the ATHOS-III trial, right? About 70% of the patients in the angiotensin II had a MAP response, which was defined as a MAP greater than 75 or an increase by 10 after initiating angiotensin II versus the placebo, where they saw it in just under 30%. So going back and looking at the angiotensin II data and these post hoc analyses, one study looked at this angiotensin I to angiotensin II ratio. And the graph on the left, so this ratio is less than 1.63, and the one on the right, you have this ratio greater than 1.63. And the blue line is representative of angiotensin II patients. So in these populations, if your angiotensin I levels were high in comparison to your angiotensin II, and you were treated with angiotensin II, you did not experience this increased risk of death, right? So that's something that catches your attention. Further looking at this, renin, it's not really feasible to measure angiotensin in one and two levels, but renin is easier to measure. And some believe that this will maybe be a surrogate for lactate or will replace lactate at some point. But this group down here was a group of patients that had elevated renin levels, but did not receive angiotensin II. And that corresponds to, or sorry, in comparison to the group that had elevated renin levels, but got angiotensin II. So the ones that did not get angiotensin II had this increased risk of death that was not seen in the other groups that either had normal to low renin levels or high renin levels that were treated with angiotensin II. This was a post hoc analysis looking at the ATHOS-3 data in just CT surgery patients. And you see 90% of the patients were responders and had blood pressure improvements. And they had a decrease in their norepinephrine equivalent dosing about to 80% or almost 80% out at 12 hours when they were started on angiotensin II. So last, I'm just gonna hit briefly on methylene blue because there are a couple of new studies that came out recently. But there's other adjuncts that other centers use. We typically go with methylene blue, hydroxyvalumin, steroids, and angiotensin II. But other folks will throw those other agents at those patients as well. And they're sick and the data's limited, so I'm not gonna blame them. So the dose, one to two mg per kg, this is one of the problems with the methylene blue data is that the dosing's all over the place. Some people put them on infusion, some people don't. Some people repeat boluses. It makes it really hard to interpret this data. Adverse effects from methylene blue are a little bit more risky than hydroxyvalumin. You got serotonin syndrome. If you have G6PD deficiency, then you might have hemolytic anemia. Hydroxyvalumin's a little bit more expensive and there's some supply issues. It's a little bit harder to get your hands on it. It also mitigates nitric oxide. It's a direct scavenger, but also decreases production. But the other mechanism is it enhances hydrogen sulfide elimination. It can interfere with some lab tests, but otherwise really doesn't have any negative adverse effects. And this is the flavor of the data that's out there in these adjuncts. This is a retrospective single center study where they go back and try a propensity match. See low numbers of patients, 66 in one arm, 22 in the other. They both raised the map, they both decreased norepinephrine equivalent doses, and it did it to similar effects. And there weren't any major differences in the outcome. This is a more recent study that came out in 2023 from the Cleveland Clinic group. Same flavor, 43 in one arm, 77 in the other. They both raised the map, they both decreased norepinephrine equivalent dosing. When you compare the two, the first study in 2022 was out at Duke and the hydroxycobalamin group was sicker in comparison to that more recent study from the Cleveland Clinic where the methylene blue group was a little bit sicker. They had a variable dosing of the methylene blue. So it makes it very hard to interpret this data, and I think it's certainly ripe for an RCT out there. So for vasoplegic syndrome, first line, I generally go with norepinephrine unless I'm dealing with RV failure as well. Second line, I'm adding vasopressin. I do that early at 0.1 mics per kilo per minute. If they're still hypotensive, then I'm looking at rescue therapies. And I think you would be justified in using any of those. And I'm really interested to kind of keep an eye on this angiotensin II data and see what comes out of that and whether we can really target some of these vasopressor therapies and adjunct therapies to the particular patients in front of us. Not use kind of a one-size-fits-all approach, but there's a lot of questions out there that persist with methylene blue and the proper dosing and timing and that kind of stuff for any of these interventions, really. So thank you for that part. All right, so next up we have Dr. Gordon. He's gonna talk about tamponade. All right, thank you guys for coming out here and for going to the beach for just a little bit longer. Today we're gonna, or next we're gonna talk about surgical tamponade and management of that. So my name's Matthew Gordon. I'm down at Temple over in Philadelphia. I have no disclosures related to this talk. And so a couple of learning objectives that we're gonna talk about. We're gonna go over the risk factors and diagnostic modalities, therapeutic options, depending on whether or not this is a short-term effusion tamponade or a longer-term process. And then focusing a little bit on some of the data around delayed pericardial effusions. It's worth noting that the internet gets better and better at predicting who you are and what you do. And when I Google anything, it is PubMed hits, home DIY, and how to restore a junkyard sailboat. And it takes me about five pages to get to anything beyond that. But if you do a clear browser and you type tamponade, you're gonna see articles like this. And this is really where the pertinence of this particular talk comes in. This is actually the supposed cause of death of Neil Armstrong semi-recently. Here's the same slide. You guys have this already. And so, okay, this is mostly just to check and see if this works, but clinically significant cardiac tamponade never occurs during banking hours, true or false? Oh, only a couple cynics here, okay. All right, all right, an optimistic crowd, I like it. So we're gonna pick up the story from Ms. Unfortunate here. 68, cabbage valve, we're up to post-op day three at this point. We heard about her initial post-op course. She had some vasoplegic shocks, some RV dysfunction, a little bit of increased chest tube output. Not terribly shocking considering the fact that she had an aortic valve done. Priswongans, catheter, and mediastinal chest tubes were all taken out on post-op day two. Pacer wires got pulled on day three. You do your evening rounds, hoping to get out of here at a reasonable hour. And you notice that there's some down-trending maps. House staff attributed this to over-diuresis. The evening LASIKs was held. They'd gotten a chest X-ray during the day, looked clear both the morning and the afternoon one. She's been afebrile for 24 hours, has a normal white count. But cultures and empiric antibiotics were started just in case because this is what always happens in the cardiac unit, at least where I am. Later in that evening, hypotension progresses to shock. She was initially fluid-responsive on two separate boluses. Not complaints of worsening shortness of breath and chest pain. EKG, low voltage, but no ischemic changes. Has some new onset AFib. Maybe we should have been a little bit more aggressive with our AFib prophylaxis. But really no ischemic changes to mention. Again, the chest X-ray is clear. Repeat, this is her third of the day so far. On exam, this is the night crew that's in there now. She's uncomfortable, she's a little bit diaphoretic. She's tachycardic. CVP is up, extremities are a little bit cool. And the working diagnosis at this point is that the day team botched it and she really does have heart failure. Diuretics get resumed promptly at that point. About 60 minutes after diuretics get administered, she has a PEA arrest. CALS protocol, pacing attempt at all without response. Emergent bedside echo during the code. Shows a large pericardial effusion. CPR gets initiated and now for the big question, what do you do? We broke the magic time, we're going, all right. We'll stop at a 14. Okay, so bedside pericardiosynthesis and a bedside sternotomy are the two leaders. A couple people want to activate the cath lab and one person wants to do CPR into the OR, perfect. So we're gonna kind of talk about what some of your different options are and what some of the different etiologies are for tamponade. And I'm actually gonna pick option, should have been F here, but G, we're gonna explore the space-time continuum. We're gonna go back and hopefully change history and prevent this unfortunate lady from coding in the first place because there are multiple points when we really could have intervened and preventing this from happening is really gonna be the mainstay of treatment here. So the first thing to ask yourself, are all effusions that you see tamponade? The answer is clearly no. Up to 85% of postoperative patients end up having some degree of pericardial effusion, be it small, medium, or large. Not all require intervention. The reoperative rate for pericardial effusion postoperatively is low, 2%, 5%, depending on the study that you're looking at. Tamponade is exceedingly rare. Depending on the study that you look at, it is reported as less than 0.1%. High end of that range is about 6%, really depends. These are not great studies. These all tend to be single center. Depending on the acuity, the type of surgery you have, the age of the patients, when the study was published is all sort of changed, but that's the rough range that we're talking about here. And this graph really just shows that. This is a single center experience looking at small, moderate, and large effusions. And what you see is that small effusions almost never really need much in the way of follow-up. These were all later effusions outside of that initial three-day window. Moderate effusions kind of split in terms of who needs to go, and then large effusions typically need to get drained. So what are some risk factors for developing a pericardial effusion? The honest answer is I can't tell you. I've read an exorbitant amount of studies trying to look at this and come up with any kind of clear answer, and the data's all over the place. Type of surgery seems to correlate to a certain degree. Other surgeries tend to be less risky than CABG. A valve may or may not, isolation may or may not be riskier than a valve plus a CABG. In general, aortic surgeries carry the highest risk of post-op effusions, which goes along with the degree of coagulopathy that you tend to see in bleeding that can occur after any of the aortic procedures. A lower Euro score and lower age actually are predictive pretty consistently in most of the trials for post-op pericardial effusion. Again, speaking of the degree of inflammation that tends to happen in younger, healthier individuals. Pre-op anticoagulant use, if you're on coubanin more so than warfarin, you have a higher risk of bleeding in pericardial effusion. An elevated pre-op PTT at baseline is predictive as well as a low hematocrit. Chronic aspirin use. Sex may or may not play a role. Older studies were pretty consistent that women were at higher risk for pericardial effusion. Most of the newer studies seem to suggest that men are at higher risk. Not much to take away from there. High chest tube output obviously has been correlated with development of pericardial effusion and ACS within the last 30 days, especially if you've got Plavix loaded. This is really one of the big things to take away and this is one of the new studies that's come out here. Atrial fibrillation and pericardial effusion. There's been several new studies that have come out that have shown a very good correlation between the presence of post-operative AFib and the development of a pericardial effusion, especially the more posterior pericardial effusion. The working diagnosis, and I'm not gonna bore you with all the biochemistry, this still gives me PTSD to this day, is that all the blood in the pericardium tends to induce an inflammatory reaction. That inflammatory reaction results in post-operative AFib and interventions that reduce your post-operative blood content or the pericardial effusion inside of it, so a la this posterior left pericardotomy, reduces your incidence of post-op AFib. So one of the things that I really want you to take away is every time you see post-operative AFib, think pericardial effusion. Doesn't mean it's gonna be there every single time, but it should be on your differential, much more so after this paper came out in the last year or so than previously. You know, we always thought of this as just sort of an unavoidable thing. It happens as part of the inflammatory pathway, the surgical connections, everything that happens when you violate the pericardium. Looks like we have another potential etiology here now. Tamponade findings. Obviously, you guys know most of the clinical stuff, right? Beck's triad, tachycardia, Pulse's paradoxus, low-voltage EKG, which we had earlier, electroalternans, you can see, in terms of Swann-Gans catheters. Depending on if your site uses those routinely for post-operative management, you're gonna see equalization of intracardiac pressure, especially the diastolic pressures. So pericardial effusions, you guys have, this has been hammered into everybody since the advent of pericardial effusions. It's not so much the volume of the pericardial effusion, it's the time that it takes to accumulate. I'm gonna push back on that a little bit in this one particular cohort, but in general, the finding holds. If you can accumulate slowly, you can get a much larger volume of fluid in that pericardial space. It will become much more compliant and stretch and distend. You don't have a lot of time in these post-op patients. If you see a large pericardial effusion, it is because of surgery, and if it's within those first three days, even within that first week, chances are you're gonna need to drain it. Most studies have consistently shown that a large pericardial effusion is gonna result in tamponade at some point and probably should be drained. If it's within that first three-day window, it absolutely needs to be addressed. Other findings that you should see, you should not have tamponade without a dilated plethoric IBC. It's incredibly sensitive. Obviously, it's not specific at all for tamponade, but it's very, very sensitive. Right-sided collapse, again, diastolic RV collapse is pretty specific, 75 to 90% depending on the study and what phase of the cardiac cycle it's occurring in. RA collapse during systole is sensitive, but not nearly as specific, but that tends to be one of your earlier signs of tamponade. For those of you that are a little bit more comfortable with ultrasound, you can use pulse-wave Doppler, and this is basically another way of doing pulses paradoxes. During inspiration and expiration, you're looking for a decrease of 25% through the mitral valve or 40% through the tricuspid valve, and that's a very, very easy way. If you wanna try and convince someone at two in the morning to come in and re-explore the chest, this is the picture that's gonna get them to come in without question. So a couple of false positives to be aware. Obviously, you can get a pleural effusion and a fat pad or the two big mimics that you wanna make sure that you're not missing there. And again, you can see that fat pad just sort of moving at the apex of the heart right there, I'm sorry, at the top of the screen. So we just spent a lot of time talking about TTEs. How useful are they in reality? The answer is they're not that great. This was another single center study. You're catching a theme here for most of the data that's out there. But really, this study was the best, but there's several others that correlate these findings in general. TTE has a false negative rate of about 21% in the early phase and about 59% overall, almost 60%. So this was a single center experience that looked at 62 patients over the last several years that had large pericardial effusions and clinical suspicion for tamponade. They looked at the TTE findings because every single patient had a TTE. Every single patient that went to the OR had a TEE. If they weren't gonna go to the OR, they had a TEE as part of their routine management and protocol, and what you saw, again, TTE has a 21% false negative rate and overall that 59% false negative rate. Almost nobody with a positive TEE finding failed to have tamponade physiology when they went into the OR and actually opened up the pericardium, whereas the TTE in and of itself, obviously, was pretty nonspecific. So it's helpful. It can help guide you in one direction or the other, but it's not something to hang your hat on completely. If you have a high clinical suspicion for tamponade, either do the TEE yourself, get to be good friends with your cardiac anesthesiologist or your cardiologist, whoever's willing to come in and do that for you in the middle of the night. And this is just another example. So this is a TEE image that shows a large clot that is invisible from a standard TTE view. So going back to our story here, how should we manage epicardial pacer wires? So these are probably one of your biggest risks of bleeding. This tends to be a late presentation of bleeding and a late presentation of tamponade. It's a rare but actually increasing incidence. Earlier studies had a bleeding rate of about 0.1% to 0.8%. And that seems to be creeping up in the more modern studies, whether or not we're doing older, sicker patients. In terms of cardiac surgery, techniques are changing. We're delaying how late we pull pacer wires. It's really not clear. And so while it's a relatively low percentage of overall bleeding complications for post-op CT surgery, when you look at late-onset bleeding or tamponade cases, it really jumps up. And it's about 12%. So one out of every 10 cases is going to be related to bleeding from a pacer wire that gets pulled. And again, this was another single-center experience. They had 11 cases from 2011 to 2016. TTE was diagnostic in less than 50% of the cases. Three out of 11, they diagnosed with tamponade. Two out of 11, they only diagnosed with an effusion. But again, the big thing to take away from this is three out of those 11 patients had a cardiac arrest. And the other eight had active bleeding at the pacing site when they went in and reoperated and went to drain the effusion. So please be very cognizant of when pacer wires get pulled. At least for us in our practice, we do not pull pacer wires after the mediastinal tubes have come out. If the mediastinals came out first, they always get cut. But again, something to think about. Obviously, everyone's going to have surgeon-specific protocols, hospital-specific protocols. But it is something to be very cognizant of. Because regardless of protocols, changes happen and these things do occur. So getting back to the case, what are some things that we could have possibly done? I just alluded to a couple of them right there. So we're at post-op day three. So now we're starting transition to that later phase of pericardial effusion and bleeding. The mediastinal chest tubes came out on day two and the wires weren't removed, unfortunately, until day three. So that right there, you know, is a red flag and a risk, especially for someone that gets a little bit hypotensive afterwards, again. And really, you've got to correlate this because, you know, these people are sick. They're complicated. They're always a little bit tachycardic. They can always be a little bit hypotensive from that post-op vasoplegic. So you really have to have an eye index suspicion. Again, I lost track of the number of low-voltage EKGs I've been handed just because it happens on a relatively common basis. But again, that should always sort of trigger something in the back of your mind. Is there an effusion? Is there tamponade? Nuance at AFib is the other thing I really, again, we really want to hammer home. When you see nuance at AFib, think blood in the pericardium, 100%. This is someone that was initially fluid responsive, got two separate boluses, got hypotensive after each of them. Again, should really, really highlight your concern. If this was just over diuresis, this should have responded to volume, instead it got progressively worse over time. And again, the physical signs that you'd expect to see, CVP elevation, cool extremities. You know, so really, if we're going to go back in time, you know, getting an echo before they coded, holding off on the diuretics, and someone tamponade really would have bought you the time to potentially intervene either with a TTE, TEE, and or surgical management there. So when it comes to actually management of these effusions, if you're concerned about tamponade, early effusions within the first 48 to 72 hours, unless proven otherwise, are related to bleeding, and those go into the OR. You can have a ruptured wall, either related to surgical complications or ischemia. You can have localized surgical bleeding. Regardless, that's a surgical problem that needs to be fixed in the OR. Draining and rapidly relieving that pressure in the cath lab can actually worsen your bleeding, can actually worsen the damage there, and at best, it's going to be a stopgap and delay definitive treatment there. Again, this is a structural problem that needs surgical repair. Late effusions and possible tamponade, if they're wire-related, again, if you pulled the wires late in the post-operative course, that's a surgical bleed that needs to go to the OR, no questions asked. If it's just a persistent post-op pericardial effusion that has been present for days and days, hasn't gotten better, continues to grow slowly, the management there actually is a little bit different, or a little bit up in the air. This is a busy slide. It's small, I know. I don't need you to know the specifics of this. All I need you to appreciate is that this was a study that looked at cardiologists and CT surgeons in a bunch of different scenarios, high-risk, low-risk tamponade, and different grades of echo, and just how varied all the responses are in terms of whether or not they would like to see just watchful waiting, more imaging, or direct intervention, be it in the OR or the catheter lab. Really, I only highlight this study because this is probably the single best one that looks at how varied the practice is from hospital to hospital. Again, this is only late pericardial effusions without abject signs of tamponade, and whether or not there is something that you can do about it. Again, if you're a surgeon, you tended to go back to the OR and drain it surgically. If you're a cardiologist, you tended to do a pericardial drain. Again, for these late cases, pericardial drains are something that's a reasonable option as a percutaneous approach, as opposed to going in for a repeat surgery. Again, so for our case, really the response after you've had a cardiac arrest is gonna depend on local expertise and availability. Within those first several days for surgical bleed, going to the cath lab really is not gonna be a viable option. It's just gonna delay definitive treatment. So your options really are gonna depend on what's available to you and what you can get done fastest. Surgical wear, depending on how savvy your staff is and how experienced you are in terms of turning your ICU into an OR and opening the chest there, versus doing CPR into the OR if you already have a crew ready to go, versus putting someone on ECMO as a bridge. Really, it's hemodynamic stability as fast as possible until you can temporize the bleeding. And so, with that, I thank you for your time and I'm gonna turn it over for the next topic. Thank you. Okay, great. Thank you so much. My only disclosure is that I'm probably one of very few neurologists here, so we're gonna take a little pivot and talk about the brain and C just now stands for cerebro instead of cardio. So let's talk about stroke after CABG or cardiac surgeries. So brief outline of what I'll discuss, incidences of stroke, classification mechanisms, recognition and response, mostly focusing on large vessel syndrome, large vessel occlusion syndrome or LVO, and then briefly reviewing the supportive evidence about hyperacute treatments. So perioperative clinical stroke or TIA after CABG is pretty common. Usually, we're thinking in the single digits percentages, the estimates really vary widely based on the study and how they determined stroke. But it is increasing over time and this is probably a combination of patients getting older, having higher baseline risk based on that, people doing more complicated procedures in a good way, more reporting and improved detection probably based on our imaging modalities. I think it's important to differentiate infarcts versus stroke. Stroke is a clinical diagnosis and when you look at studies that actually use serial neurology assessments and imaging, this number more than doubles. And if you really look at clinically occult but radiographically proven infarcts, this is dramatically more prevalent. So my buddy at Penn did a study looking at this and they did serial MRIs without any clinical symptoms and this was greater than 50% after cardiac surgery. So this is huge. And I think in neurology, it's location, location, location. It's like real estate, right? So you can have a stroke the size of a lentil and in a certain part of your brain, you could be hemiplegic, aphasic and then in another location, you'd have no symptoms. So really these clinically sort of quote unquote silent infarcts, it's really luck. And so I think it's important to also look at this data. But it's really about reporting, right? The surgeons are not gonna want to report complications of their surgery. So the studies are pretty variable. Why do we care? Stroke increases hospital length, cost of stay and over 10 times the hospital mortality. And in addition to actual stroke and stroke symptoms and disability, it also increases risk of other neurologic complications including seizures, delirium, encephalopathy. I think these statistics are pretty compelling. Patients consistently report a worse quality of life after stroke than even heart attack. And more than 80% of patients would prioritize stroke freedom over longevity. People are afraid of stroke and I think that's really appropriate. Okay, so mechanisms and timing. I always like to joke with my cardiology colleagues that they have it easy, it's all athero. Maybe a little scat, maybe a little vasospasm but mostly athero, right? For stroke, it's different. It's really pretty split evenly. So you can see there's definitely plenty of large vessel athero but a significant amount of cardio embolism, small vessel disease and actually a large proportion of patients even after a comprehensive stroke workup are still cryptogenic or embolic stroke of unknown source, ESUS is what we say. And some have competing high-risk mechanisms. It doesn't have to be one or the other. The exact percentages vary depending on what the population you're looking at. And the reason it matters is that different mechanisms are actually nowadays treated differently. This wasn't the case even 15, 20 years ago, have different prognosis and rates of recurrence. Let's talk specifically about stroke after CABG or cardiac surgery. The vast majority in this case are actually due to cardio aortic thromboembolism as you can imagine. So a lot of this is aortic manipulation. These are vascular paths. They have tons of athero everywhere. So you can see they're doing cross clamping. That's a gross arch that's gonna definitely hold risk, high risk for embolism to the brain. And AFib, as we already talked about and we'll talk about more, this can either be early AFib or delayed AFib. AFib, so this is our nemesis in stroke. I think it was Dr. Gordon that said, you hear post-op AFib and you think pericardial fusion. I would argue you hear post-op AFib and you think stroke. So you look at this and I think we all know it's very common but these numbers are staggering, right? So after CABG and valve replacement, which happens all the time, more than 60% have post-op AFib, right? So that is a huge number. This is probably one of those one plus one equals five situations where it can be some pre-existing myocardial changes, pre-morbid existing risks and then also that perioperative predisposition to arrhythmia that sort of arrhythmic milieu. There are some risks that you can look at to say, hey, maybe I should be more suspicious of AFib but I would argue with these numbers of 60%, you should be nervous about all of your patients having post-op AFib and have a very low suspicion in all of your patients. Sorry, low threshold, high suspicion. All right, so one other mechanism that's pretty common after heart surgery that's not necessarily one that's otherwise common is border zone or watershed infarcts. So this is in about up to a third of patients. There are two kinds of border zones, cortical or internal. The cortical you can see in blue there and usually related to thromboembolism if it has that appearance and the internal border zone infarcts which is really the bottom red pictures are mostly related to poor perfusion to the brain. It's a little more complicated than that. I won't get into it in the interest of time but even if you have an embolic shower, if you have hypoperfusion also, it can significantly impair the clearance of emboli, that washout and either increase stroke risk or severity so you can have both and that sucks. All right, let's talk about stroke recognition. We're gonna get back to our patient. Spoiler alert, you probably knew from her name she was gonna have a stroke. So after returning from surgical exploration, the bedside nurse noted she was not moving the right side as much as the left and not responding to questions. The nurse immediately notified the primary team who appropriately activated a stroke code. She was noted to be an AFib with RVR on the tele. What does the stroke team need to know when we get there? We need to know time lasts well. This is not the time you notice the symptoms. This is the time the patient was last actually witnessed by anyone to be at their neurologic baseline. This is oftentimes preoperative and that may have been many hours ago. Does the patient have decision-making capacity? Are they on an antithrombotic therapy? Many of these post-op patients will not be on anticoagulants but you may. What is their blood pressure now and actually in the recent past? And there's some other potentially relevant data based on your patient, renal function, their ejection fraction. These can matter for advanced imaging. It can affect bolus timing for contrast on CT perfusion and CTA. And their baseline functional status. Is it worth pursuing an aggressive intervention for this person who is having a stroke? Keep in mind, most post-surgical patients, as you probably already know, will not be candidates for thrombolysis. The other emergent intervention we often do for stroke, TPA, alteplase, nowadays tenecteplase. One caveat is we're doing more endovascular transcatheter valve replacements. Those you might consider on a case-by-case basis based on interdisciplinary discussion with the surgeons, of course. All right, so I think we really need to focus on large vessel occlusion, LVO. This is hot right now because basically we have new data, which I'll show you, to support doing these interventions in stroke patients even many, many hours out and in the post-operative population. So generally speaking, LVO refers to an occlusion in the internal carotid artery or the proximal branches of the MCA, middle cerebral artery, M1, M2, then the vertebrobasilar arteries. These are the biggest blood vessels in the brain. Posterior circulation, vertebrobasilar, not as well studied, but it has such high morbidity and mortality strokes in that area that we really treat them, if anything, more aggressively, even though the data is a little bit less robust. We're starting to study more medium vessel occlusions. Some people call these MEVO, just with more limited data, but they can also be very morbid. Keep in mind that from traditional definitions, only about 20% of stroke patients will have an LVO. We wish there was more. We wish there were more patients we could emergently treat, but this is gonna be the minority of patients. That number probably doubles if you include more distal occlusions, which we are now starting to treat also with rhombectomy. How can you suggest that maybe your patient is going to have an LVO, just based on their clinical exam? Cortical symptoms are extremely suggestive, so this would include gaze deviation. If you have a problem with the frontal eye fields, dominant hemisphere, usually the left, can cause aphasia, either expressive or receptive. The non-dominant, usually right hemisphere, can cause serious neglect. They don't even recognize they're having a problem, which can really complicate their exam. And then an occipital or visual field cut, homonymous hemianopsia, they can't see anything on half of space in both eyes. You can imagine if the patient's intubated or sedated, they really, a lot of these are gonna be very difficult to assess, and that is a challenge. This is just to show you there are a number of scales or scores that are looking to predict LVO. These are largely used by EMS for pre-hospital triage to take patients to the appropriate stroke center, but there's no favorite, there's no best, so a lot of these are just used interchangeably. We came up with our own score at NYU based on a retrospective study of patients we had treated over many years for thrombectomy to try to capture as many as we could, which includes weakness, face, arm, aphasia, neglect, and obviously prioritizes cortical symptoms to help our staff decide when to call a code or consult. Code mobilizes lots of resources and is looking for patients for hyperacute treatment. All right, so back to Ms. Unfortunate. We're looking at our stroke code exam. She has a right gaze deviation, no blink to threat on the right side. She's nonverbal, doesn't even follow simple commands, has a right facial droop, and is plagiac on the right despite noxious. So what would we need to do before we even consider an intervention? She needs a non-contrast head CT. She needs usually advanced imaging, a CT perfusion or CTA head and neck. Some places will do MRI, but most places probably not. If there is no occlusion, there's no opportunity for intervention. So we are really hoping for an occlusion that we can treat because, again, these patients are not candidate for lysis usually. Risk-benefit discussion. Follow your local consent process. You should know what it is. At NYU, we do not require formal written consent because this is standard of care at this point, and I'll show you the data that supports that. And it is an emergency, just like a trauma. Interdisciplinary discussion with the primary surgical team, and then lots of resources and people need to be mobilized. Anesthesia, the neurointerventional team, to be able to do these at any hour. So this is a non-contrast head CT. Even if you don't know what you're looking at in the brain, you can see there is actually gaze deviation, and we can often see this on CTs. So that's one sort of indicating factor. And then there's subtle changes, early ischemic changes. There's loss of gray-white differentiation on the left versus the right. Maybe it doesn't show up that great on the screen, but basically we're looking for no hemorrhage and no major stroke already identified. If there is a clearly demarcated stroke already on the non-con CT, you're probably too late. That infarct has probably completed and there may not be something to save. Nonetheless, we get advanced imaging, and we'll look at that in a little bit. So let me briefly review the data supporting endovascular therapy. These studies are sort of a collection that I compiled of the randomized controlled studies that initially came out in probably 2015, 2016. And these are all over varying inclusion criteria, populations across the world. And all of them are looking at endovascular intervention, so mechanical thrombectomy, plus best medical care versus best medical care alone. Most of the patients in both groups received TPA if they were eligible, I'm talking 80, 90%. So these are not studies comparing endovascular intervention versus TPA. That's a common misconception. Most patients, the vast majority in both groups receive TPA. So we're talking endovascular plus TPA versus TPA alone. And you can see if you look at a lot of studies that despite this variability, these number needed to treats are impressive by any standards. Look at these single digits across the board. So nonetheless, the AHA came out almost immediately with updated guidelines, suggesting class 1A recommendations to do thrombectomy in anterior circulation LVOs up to six hours. Time continues to be vitally important on an individual patient basis. This is just a meta-analysis in the Lancet that just shows you that the pre-specified subgroups don't change anything. So that doesn't matter if the patient's older or younger, if their infarct is large or small, if they got TPA or not. None of that seemed to matter. It's still dramatically favored intervention in all the subgroups. So we said, all right, cool, we can do this up to six hours. Can we extend the window to be even longer? Because as you know, a lot of these patients aren't identified until post-operative and then they wean the sedation and they say, oh, actually this person's not really moving the right side. So the answer is actually yes. We looked at this in Don and Diffuse, these two trials that basically supported that in the right patient with a favorable perfusion imaging. Even up to 24 hours, these patients dramatically benefit from thrombectomy. So this is really important for the post-op population where this is your only treatment option. So the AHA guidelines, again, updated and still class 1A recommendation between six and 16 hours, 2A between 16 and 24 hours, again, with eligible imaging. So as I was suggesting, this is being studied right now, more distal clots, because we want to just help as many people as we can, and you can have an M2 or an M3 occlusion that causes significant morbidity and complete aphasia. So there's 2B level of evidence and recommendations supporting more distal thrombectomies. So this is our patient scans, her CTA on the left there, and you can see, even if you don't, does this pointer work, that this is the occlusion here in the proximal M1 on the left, and that there's a huge gap here in perfusion, and her brain is very, very sad on that side. This is the CT perfusion. You can see that the pink here is the core infarction. That's already done by the time you even identify the symptoms and get the scan, call the stroke code, but the green here is the tissue at risk, the perfusion mismatch, or sorry, this is the penumbra. So the mismatch is the difference between the pink and the green, and there's actually pretty large territory of brain that could be impacted if you don't recanalize quickly. So we took her to the Andrew suite, and this is pre-thrombectomy. Again, you can see, you kind of have to be familiar with what you're looking at, but there is an occlusion here, and this whole part of brain should be getting perfusion. After thrombectomy, it is. So you can see this restoration of blood flow to essentially half her brain, this entire left hemisphere is now being perfused appropriately. So clinically, her exam started this way. After thrombectomy, no gaze deviation, full visual fields, speaking normally, following commands, does still have a mild right facial droop, some mild right-sided weakness that's more pronounced distally, but went from an NH stroke scale of 25 to three. You can really have miracles here. Radiographically, what would this look like? This is what her brain scan, her MRI, could have looked like. This is a huge MCA infarct. That's basically most of the left side of her brain. That is not what her scan looks like, thank goodness, because we saved her. And so this is what her actual MRI looked like, which is just actually a small core infarct deep here in the structures that were already infarcted by the time the code was called. It actually matches up really well. If you remember the perfusion imaging with the pink area, that was already done. That wasn't salvageable, but anything we could have saved in that green, we did. All of her cortical areas are spared. So we talked about risk benefits with the surgeons and decided, based on the post-op AFib, to start new apixaban, Eliquis. She actually was able to ambulate with the walker and feed herself and discharge to acute rehab and was very grateful for the care. So in summary, perioperative stroke after a CABG is common. Most studies probably dramatically underestimate the incidence, and clinically, quote-unquote, silent infarcts may approach 50% and keep in mind, these are not really clinically silent. They might have serious impacts on delirium, cause seizures, delay hospital discharge, all of these things. Mechanisms can vary by timing, but primarily in the post-op population, these are embolic and or related to hypoperfusion, and post-op AFib is so common. We're talking like up to 50% or more in some patients. You should have a very low threshold to call a stroke code for any focal neurologic symptom post-op. We can do much more than we used to be able to, even five, 10 years ago, and this is because there's now extensive evidence growing every week that supports the efficacy of endovascular intervention for acute ischemic strokes with a large vessel occlusion, including post-operatively, and in many cases, that is the only intervention that we can offer people we know are actively having a huge stroke. Thank you.

stroke after cardiac surgery

large vessel occlusion syndrome

perioperative stroke

TIA after cardiac surgery

infarcts and stroke

clinical diagnosis

cardio-aortic thromboembolism

recognizing stroke symptoms

thrombectomy for LVO