So, first of all, welcome to this beautiful city of San Francisco. I was going to say warm as well, but Mike made me sure that I will not lie on this podium. And I will be talking about how to diagnose and imagine sepsis cardiomyopathy. My name is Siddharth Dugar. I trained in critical care medicine at Cleveland Clinic, and I'm also the director of point of care ultrasound. Disclosure, I have a chest sinusoid grant to study the factors that lead to DVT in COVID-19 patients, so nothing related to this topic. So we will be going over the definition and what are the different imaging modalities and where does the future of sepsis cardiomyopathy lies. So again, everybody in this room knows about sepsis. It causes dysregulated inflammatory response, affects every organ in the body. The center of that is the cardiovascular system. When we talk about sepsis cardiomyopathy, what we are trying to do is parse out the cardiac dysfunction from the cardiovascular dysfunction, but they are so intertwined. And as we go through the slide, you will see that it's very challenging to figure out, is it just cardiac dysfunction or cardiovascular dysfunction? So definition, as Dr. Parker already has gone through it, it's acute or acute on chronic. It happens very early in the course of sepsis. It usually causes global dysfunction. You have to make sure that you rule out any ischemic causes, and it's reversible. It can affect systolic. It can impair diastolic function of the ventricles, and it can affect the right or the left ventricle. So when we look at the entire prevalence of sepsis cardiomyopathy, we see that in the first one to three days of sepsis, the incidence ranges from 40 to 60%. This is the seminal paper. Anybody who is interested in sepsis cardiomyopathy has to read this paper by Dr. Parker et al. So if I make a mistake, you can always ask Dr. Parker to explain it. So again, I put QR code, so if anybody's interested in reading the paper in detail, just scan the QR code. So the seminal paper, the first plot, looked at 20 septic patients, and they traced the LV systolic function by measuring LVEF throughout their course of sepsis. So the important thing is that the reduction in most of the non-survivors did not have any changes in their LV systolic function as measured by LVEF, but there were 13 survivors, and 10 of those patients actually had reduction in their LVEF by day one or two. That's when you will see most of the impact of sepsis cardiomyopathy. The other plot here is a study that was done by Nathanson et al., published in 1987. So again, a very early study, but a really good study where they took 10 dogs, and they induced sepsis by peritoneal sepsis, and they traced the LVEF over the course of minus seven days to plus 10 days. And they found, again, the same thing. The most of the reduction that we see in sepsis cardiomyopathy happens very early in the course. So if your patient has sepsis and you scan them at four days, you may miss the diagnosis of sepsis. So it's very early, and if they survive, most of the time their systolic function will reverse. There are other modalities too, biomarkers. We have CK, troponin, BNP, and I've cited the best paper I've found on those topics. There are newer imaging modalities as well. MRI, MRI can look at actual myocardial injury, and PV loops, or pressure volume loops of the LV. Again, the most gold standard, non-load dependent measurement there is. But these are still in research realm. There is not a great study that I can cite which has been done in sepsis patient, but these are some of the tools that can be used to truly define what patients with myocardial injury or sepsis cardiomyopathy. Again, going back to the same paper, we will be talking about echocardiogram because it is kind of the whole and soul of how we diagnose sepsis cardiomyopathy. Going to the same paper published in 1984, interesting thing is most of the survivors actually had reduced LVEF at the time that you diagnosed sepsis cardiomyopathy. As you trace them, by date seven to 10, the cardiac function returned to normal. But the important thing here is even though they had depression in their cardiac function, the LV was able to dilate and increase the filling pressures and was able to maintain the stroke volume, which is what we care about. So if you trace the stroke volume of all the patients and survivor versus non-survivor, it was just a little bit low, but it was still within the normal range. So the best way to interpret, or the way I interpret it is even though they have depression in their cardiac function, they were able to increase their ventricular filling and maintain their cardiac output, and those patients had the best outcome. Or reverse-wise, we can say patients with hyperdynamic circulation or with no variability in their LVEF during their sepsis had the worst outcome. So the first one is LVEF, again, we have talked about it, it's load-dependent. Most of the times when we do LV systolic function assessment, we use eyeball in the ICU, which is not as sensitive. Again, we cannot define if it is 45% or 35% using eyeball testing. So if you are going to use LVEF to determine if somebody has sepsis cardiomyopathy, make sure that you do biplane or monoplane Simpson method. The definition we have is LVEF less than 40% to 50%, again, different studies have used different cutoff. I think 45% makes more sense to me. And they recently added an acute reduction in LVEF by 10%. Again, people who have cardiac dysfunction can also develop sepsis cardiomyopathy. And this is a study that has looked at the systematic review of how LVEF affects outcome in patient with sepsis. Strain, so you will, as going forward, a lot of the studies that will look at sepsis cardiomyopathy will be using strain. What happens is when ultrasound interacts with the myocardial tissue, they generate small dots in the myocardium. And all we are looking at is how much does the distance between those two spots decrease during systole. Since we are looking at reduction, the value is negative. The more negative it is, the worse it is. So abnormal is when it is more than minus 17%. So 10 is bad. In the image, you can see minus 3.8, really bad. It is less load dependent. It is very good at detecting sepsis cardiomyopathy. So if you look at the paper on the QR code, they found that when we used LVEF, we found the incidence to be 22%. When we used strain, the incidence of sepsis cardiomyopathy was 70%. Reason being, if patients have low afterload, they are able to increase their LV systolic function if you use LVEF or conventional methods. But strain is able to detect the myocardial dysfunction much better. Also, when they used strain, they found that people, when they found a strain to be abnormal, even when they had normal LVEF, two days later, they did develop a reduction in their LVEF. So good at detecting and also predicting sepsis cardiomyopathy. Next is MAPC or MAPSV. This is my personal favorite, just because it is very easy to get. Anybody who does echo in critically ill patients will see that it's very challenging to get images. All we are looking here is the distance that is traveled either by the lateral or septal mitral annulus or the myocardial velocity. There are not enough studies for me to tell you what is normal, what is abnormal. But these are some of the values that I have found. The interesting part of this is it's very easy to obtain. Even if your image is not great, you can obtain these values. And if you read the first QR code paper, they found that it is afterload independent, which is one of the biggest challenges that we face when we look at sepsis cardiomyopathy. And if you have resuscitated the patient enough, you may find that it becomes even preload independent. So a very good measurement that can be obtained very easily on a critically ill patient. The second paper is a systematic review done in 2017 that looked at the impact of MAPC or MAPC on outcomes. Diastolic impairment, the total incidence of diastolic impairment is close to 48%, pretty high in our sepsis and septic group. Again, what we look at is the early relaxation myocardial velocity at lateral and mitral annulus. Some use lateral annulus. Most people use average E prime value. And a septal E prime that is less than 7, or a lateral E prime that is less than 10, is what we use to define presence of impaired relaxation. When we start using E or E prime, it is a non-invasive marker of what the feeling pressure is. Again, if the myocardial is impaired, it will relax more slowly. You will require higher feeling pressures to maintain cardiac output. And that pressure will be felt by all the organs. And that's why it makes sense that if you have diastolic dysfunction, your outcomes will be poorer. Next slide. This one is, okay, perfect. So systolic impairment of right ventricle, again, a lot of studies recently have shown that RV dysfunction actually, or systolic dysfunction actually, is associated with poor outcome. The challenge there is most of the causes of RV dysfunction is LV dysfunction or diastolic dysfunction. So is it truly RV dysfunction, or are we using better measurements that can define normal from abnormal? Total incidence, 35%. That is a systematic review that was published in CHEST, showed that RV dysfunction using either TAPC or RVS prime was associated with poor outcome. The challenge with this measurement, as with any, is that it's very angle-dependent. It is measuring a velocity or distance at a single value. So if you have global dysfunction or if you have radial dysfunction, these values may not be able to give you the full picture of RV dysfunction. That's why my personal favorite is RV fractional area change. It is similar to what we do with LV or LVEF. Similar to what we do with LVEF, where we are looking at both the longitudinal function and radial function. The challenge is you have to get an image that is of such good quality to be able to measure the RV fractional area change. The normal value is less than 35%. Similar to what we do with LV, we also have a strain of the RV. We can measure free wall strain on the right side. Again, same thing, negative value. The more negative it is, the better your RV's systolic function. And here, the value is minus 21%, and for the LV, it was minus 17%. This is a paper that was published by Dr. Lance Pytel that looked at use of this measurement, particularly RV fractional area change to define RV dysfunction, and how it was associated with worse outcome in sepsis and septic shock. So where does the future lie? Again, we have talked about it. As you can see, most of the measurements we are talking about are either very load dependent or somewhat load dependent. We can talk about more load-independent variables, like using PV loops, afterload-related cardiac performance, where we try to see what they are predicted versus true LV cardiac output is, and that is one measurement that becomes independent of the afterload. A lot of interest is in ventricular arterial coupling. The best way to understand is how much synergy is between the ventricles and the arterial system. The more synergistic they are, the better they function. And that can be one of the things that we can intervene on to improve outcomes. Myocardial performance index, again, very challenging to do, where we try to see of all the time that you have in systole or diastole, how much time are you actually spending in true systolic versus isovolumetric contraction or isovolumetric relaxation. The easiest thing I can think about is, in SACM, if you spend more time attending conferences or networking, you are doing a better job. Severity of – so we talked about like less than 50 percent, but if an LVEF is 20 percent versus 48 percent, the outcomes are different. We haven't looked at severity. We are still defining it. And again, like in the future, I think we should be talking about severity, 10 percent reduction, 20 percent reduction, and that will help us what truly affects outcome. Interactive model. Most of the studies that we have looked at have looked at LV systolic dysfunction away from – RV systolic dysfunction away from diastolic dysfunction. But patients don't come with one problem only. They have multitudes of cardiac dysfunction. So we should be looking at more interactive model. What if patient has RV and LV dysfunction? Is their outcome worse than just LV dysfunction? Automation. Again, we talked about a lot of measurements. Those are very, very challenging to do in critically ill patients. And again, if you ask me to do fractional area change every five minutes on my patient, I will say like, you know, like, you know, maybe I will take a day off that day. But automation may help you where the machine is able to give you RV fractional area change and all our job then becomes is to decide on the intervention. And I will leave you with one last question. So this is a paper that we just published in CHEST where we also showed the same thing. Hyperdynamic circulation is bad. But we showed a U-shaped curve in all sepsis, septic shock, and severe septic shock. We have just discussed about reduction in cardiac function. We haven't talked about the hyperdynamic circulation. And it may all be afterload. But is there a phenotype where there is increase in the myocardial activity because of inflammatory situation? With that, I will end.

sepsis cardiomyopathy

diagnosis

imaging

echocardiogram

cardiac function