Welcome everyone to this session and the final session of today's masterclass of Saving the Kidneys. My name is Raj Basu. I'm your moderator and I'm going to be talking to you about something that I consider a kidney injury parallel to the blood gas. Just a few consulting relationships to disclose to you. I'm very grateful to the Society for allowing us to put this session together and hopefully you found it very interesting and contemporary for your practice. I wanted to start by asking you a simple question about how do we drive? If you think over time we've evolved slightly and we could say we've gone to this era from classic cars to more modern cars and you would say to yourself well that's not what my dashboard looks like but I want you to keep this tucked in the back of your mind because as you've heard today already we're a little bit behind in the field of AKI and rear replacement therapy. We should be thinking about this idea of targeted approaches to AKI. Unfortunately we've been limited somewhat in our diagnostics. As the other speakers have alluded to we really have not gone very far in both integrating biomarkers that are new or thinking about biomarkers that are new in a new way or even thinking about creatinine in a new way. We've relied on it for decades to do almost everything and that's not the fault of creatinine even though we like to blame creatinine for the idea that it's a flawed marker. It's not really we're just asking it to do too much and doing more than it actually could. So creatinine is a problem but it's not the only problem. A lot of the problem is that we're asking it to be a specific specific marker for the type of injury, etiology, location, and severity. That may be our own problem. In reality if you look at the data about this idea of replacing creatinine the goal that others have alluded to of finding a quote renal troponin, there's a two-dimensional quest for the perfect biomarker. One that labels the biomarker versus the injury being AKI, a static versus static mathematical equation. And over two decades of data have indicated that a lot of people have a lot of information. Nearly at this point if you do a 2022 analysis of PubMed per year there's probably 400 ish publications about AKI biomarkers reported across populations including the full age spectrum of patients in critical care and in the hospital population. There are biomarkers as you've heard some of them that are talked about in terms of early detection and prognosis based in both urine and serum in differential diagnosis and here are smattering of them. This is certainly not an exhaustive list of the biomarkers that have been discussed. The issue is what are we doing with the information that comes out? What do we do with these biomarkers that have been reported? This isn't a very recent review of the literature but it's pretty similar to the ones that have been done recently is that the papers that are published look a lot like this. Tomes and tomes of numbers and values related to the definition of AKI and the predictive value of these biomarkers in isolation. What we end up getting from this is kind of unknown because what we're trying to find is actionable information but what we're doing is population level data study. What we're doing is epidemiologic surveys and the efficacy for predictive variability of these biomarkers and maybe not related to the specific patient. What we want to do as practitioners is to identify individuals, to identify what biomarkers can tell us versus static endpoints, maybe not even about disease severity. Because what we've done in reality is we have created a situation where the biomarkers have been studied for 20 some years but they're at war with each other and in case you think I'm wrong, any new publication will have data that looks like this and this is not even new but every biomarker that comes up has to be compared to the others and specifically with the AUC value or area under the curve which is in and of itself a predicate based on Pearl Harbor and a war metric but it's compared to each other and that's what we determine as functional as a biomarker. AUC is for prediction of AKI at a stage 3 or ultimately something like receiving replacement therapy. I'm not sure that that's the way the right way to look at it because it's based on a singular construct that the endpoint, the Y, is the same. So take a look at these two. Many of you would look at them and say no thanks, I don't care, it doesn't matter but it does matter that a banana that looks this way, a brown yucky looking banana, it's not always the same reason and perhaps one you found underneath your car because your, you know, your grade school kid decided to throw the banana away from lunch and left it underneath your car and the other you put in your fridge. Well they're different in their consistency, they're different in their color, in the reason that they're dark and that makes it different overall. It's not that they're both brown for the same reason but yet we apply AKI the same way, that all AKI is the same, all the biomarkers work the same, they're all telling us these things and it's not true because if you had a banana you may not give the same one to your child versus your mother-in-law, correct? So we shouldn't be thinking about AKI that way either. The AKI phenotype is complex and therefore we should have a similarly complex construct and way we use to study AKI. We have those ways. So if you think about the complex AKI phenotype, it does matter to know where the AKI is occurring. It's not simply okay to say, oh the creatinine is elevated so there's injury in the kidney. We would never tolerate this for other disease processes. If you had an elevated white blood cell count and you told an oncologist, I think this patient has cancer, what would the response be? Be ridiculous. But we do that for kidney injury. But we know that AKI can occur anywhere in the proximal tubule, distal tubular structure, the mesenchyme. It can happen because of different biopathophysiologic mechanisms that happen in different regions of the kidney. We know that AKI varies based on timing and severity. So data that we have even from children, which is this study, demonstrates remarkable phenotypic difference of the outcome of children based on what their AKI temporal profile is. We also know mechanistically that we know very little actually about what's happening. So yes, the creatinine is elevated. Or yes, there's oliguria. But what does that actually tell us about the mechanistics of injury? These independent papers have discussed, well biomarkers can tell us about the phase of injury or what's happening based on the extent of injury or the repair of injury. We can determine whether the mechanism is a tubular injury or cell cycle arrest type of injury. Perhaps the genetic dysregulation on different biological processes affects what happens in the kidney and yet we only have one readout for it or two readouts. That's probably not okay. In fact, if you look at ischemic models of AKI, it'll tell you, the models will tell you, tremendous variation heterogeneity in the idea of what genetics are being dysregulated in the form of AKI. Yet we just rely on one marker or two markers checked at a single point in time. But our paradigm has so many gaps in it. Maybe we can determine, establish a injury. Maybe we have an idea to do first clinical recognition. But we don't really even know much at all about how to predict recovery, how to think about management during the progression, during the recovery, how to prevent future disease based on a biomarker profile. We just don't know those things. It's because we're not even thinking about it. We're not thinking about the integration of biomarkers as they apply to our actual management of care. That's why phenotyping AKI is relevant. That's why these kind of constructs can help us. Because the truth is, we use a three-dimensional, and one would argue maybe four-dimensional, axes of phenotypic information to help us. If you only have the disease, you only have data, if you only have the disease, or if you only have the biomarker, but if you can marry those two, perhaps you have knowledge. But if you know it over time, that's wisdom. Which is why old experience is hard to replace, and it's the difference between a young trainee and a very seasoned attending. May not be actual knowledge or data, but a lot of it is wisdom developed over time. And the biomarkers do have information about this, and I alluded to this before, about how they change over time and what they tell us. We also know, based on several recent data sets, that if you measured biomarkers as they change over time, they may be able to inform clinical practices. And this was a series of cases we put together that helped demonstrate the utility of biomarker series checks over time and how they affect fluid management. Other people have discussed different biomarkers in the aspect of how these things change in response to an injury, such as surgery, or exposure, such as to antibiotics, or for repeated use of NSAIDs. These are things that we should perhaps be thinking about as we navigate how to even use something like Zosyn that you see in your bottom right corner. Perhaps the biomarker profile will tell us in real time how to manage those things. And if we think about management over time and how combinations of information yield phenotype and then allow us to identify a target, think about respiratory failure. You would never just take a singular physical exam or singular x-ray finding and use it in isolation to give you an idea of, nah, there's no disease, so I'll see you tomorrow. Or yes, there is disease, so you get a piece of plastic down your throat. You would never do that. But we marry pH, hemodynamics, blood gas findings, secretions, a physical exam, to give us an idea of what the patient is actually going through. Context matters. The phenotype matters. And we do this over time. And I would tell you that a blood gas is actually our best window over time to tell us how a patient's doing, or a patient's responding, or they're agnostically what to do for a patient. And I would ask you, do we do that for AKI? Have we tried that for AKI? Perhaps not. Because we should be asking ourselves not how we drive right now, but how should we drive? Perhaps the combination of GPS and modern technology and Waze and other kind of adjuncts help us chart a course so we can prophylactically drive, so we can preemptively drive, so we think about our management in real time. And so I don't think replacing creatinine, or what I call troponin worship, is the right thing. It's very monocular. We have envy, and when I say we, it's in the Society for Critical Care and Nephrology groups, of the cardiologists. Their ability to escalate their diagnostic capability, because now they have lots of troponin isoforms and imaging data and cath and newer tests, because you would never tell a cardiologist, well, the LV shortening fraction is this, and so this is what the diagnosis is. No, that's not true. But we have envy for that, because all we've had is creatinine. But the heart is not necessarily to be worshipped. It's a muscle. The kidneys do so much, and this is part of the argument to why we should be focused on marrying complex biomarkers and complex phenotypic analysis into this idea of how do we think about kidney injury. It is deserving of the attention. So is this the right way to drive? I don't know. I will tell you no, because I think what we've been chasing is the wrong model of care. We've been thinking about kidney injury in the sense of, it's just like acute coronary syndrome, or MI, that we have the biomarker that will tell us that this is happening and then we can act. I don't think that's the way it works. Perhaps we should think about a different analogy, and if you think of one that may be applicable, maybe it's ARDS and acute lung injury. There are a lot of similarities in the way that ARDS is manifest biologically, at the cellular level and at the physical level, to acute kidney injury. And one only need look at acute respiratory distress syndrome and the biopathophysiology to understand. To understand, okay, the management of personalized ventilation in ARDS, and the way we think about lung compliance, and the way we think about ventilatory strategy, perhaps is very similar to how we think about kidney injury, and how we think about CRT use, and how we think about the complexity of the pathology that's occurring, and not just ischemic injury. Because that may not just be the singular story, and we appreciate more and more from a biological level how it's so complex. So the new directive is not that acute coronary syndrome is the goal, perhaps respiratory distress is the goal. Maybe that's the model to recapitulate. Lots of inputs, most commonly sepsis, progressive, just like AKI, multifactorial with direct and indirect or injury manifestations, just like AKI. It's not acute coronary syndrome that replicates AKI, it's acute lung injury. So if we think about the X and the Y and Z, the wisdom model, versus just a renal troponin, how do we think about this? I would say we have what we need. We have a marker of homeostasis, we have a marker of tubular epithelial function, which is NGAL. We have a marker filtration. Oh, by the way, creatinine is a good one. Creatinine is a good marker. We shouldn't be throwing it away. The functional reserve in the kidney or furosemide stress test. Compensation in the system, which could be percent fluid overload or one of the newer markers called CCL 14. Stress in the marker, like a lactate, this would be the TMP2-IGFBP7 marker or the FDA approved nephrocheck marker. We can marry them all together. So just like in a blood gas, perhaps there's an AKI biomarker composite that tells us the same thing about homeostasis, solute transfer, tubular epithelial function, reserve, compensation, and system stress. Maybe if we thought about a composite tested over time, we would have the same way of thinking about AKI that we do with ALI in the use of a blood gas. Acidosis is a manifestation of too much consumption and or consumption that exceeds delivery. Oliguria is the same way in the kidney. Too much consumption, not enough delivery. The response to acidosis is multidimensional, just like the response to oliguria. They don't mean anything in isolation. They don't mean anything when you just say a patient's oliguric. What about the other aspects of the body? If you're acidotic but everything else is okay, what do you do with that? Hypoxemia is impaired filtration at the level of the alveolus. No gas gets across to the pulmonary capillary bed. That's the same thing as elevated creatinine. It may be the kidney's conservation mechanism, for instance, the hypoxic pulmonary vasoconstriction equivalent. It's just impaired filtration. So when we have response, when we have a decreased arterial oxygen tension, our body increases perfusion. When we have increased serum creatinine, what do we do? What do they mean in isolation? What are they? Static. We don't do that for acute lung injury. Why do we do it for acute kidney injury? If we have hypercarbia, impaired gas sweep, a CO2 elevation, that tells you something about pulmonary tubular function. An elevated tubular damage biomarker tells you the same thing, that maybe we can't handle fluid. Maybe fluid needs to be removed somehow. If you marry these things together, what does it tell you? So perhaps when you think about it also, I made this argument of the system under stress. Lactate in and of itself means nothing, but when you look at a high lactate in the context of other things, you can determine if the patient's under stress. Just like a nephrocheck or IGF, TMP2-IGFPP7 can tell you. Is the response to a lactate tell you something? Does a response to a nephrocheck tell you something? A response to a TMP2-IGFPP7? That is how the PREV-ACI bundle worked. Dr. Sauerbach and his colleagues that have reported numerous publications from their German study looking at cardiac surgery patients bundled to AKI management care based on a biomarker stratification. So if you integrated this, wouldn't it be fascinating that you could answer this question? This critically ill patient who's not making urine, what do they need? More fluids or diuretic? Do they need a vasopressor or something else? Isn't that a very contrarian question that we constantly answer? Perhaps the ABC-AKI could give you a sense of what actually makes sense because you have a multidimensional longitudinal approach to AKI. This Dunn-RRT study, which is a systematic review of looking at biomarkers and the liberation from renal replacement therapy, is crucial. Dr. Conard talked about de-resuscitation. How do we do that? How do we take someone's CRT off? It's not that hard, you can unplug it. What you really want to know is if they can come off and stay off. But maybe the AKI blood gas, the ABC, can give us the information about someone's stability to come off. Maybe they're not under system stress, maybe they have enough reserve, maybe their tubular epithelial functions intact. Those things would give you more confidence and they should because you're managing AKI in a multidimensional longitudinal way. So during this master class you've heard a lot about the integration evolution of AKI and RRT. What we're trying to tell you is that we're turning data into wisdom. It's taking time and what it's actually taking is a multi-professional, multi-disciplinary approach that incorporates not just physicians, that incorporates all of us to turn our knowledge into real-time actionable information. Because when we put it together, we should be flying not driving. We should be thinking about how do we actually move things forward and the way to do that is dynamic management. And we've heard a lot about that today and I hope the master class gives you seeds that will germinate into actual work. We look forward to working with you and I'd like to thank the Society for allowing us to put this together and some acknowledgments at the end. Thank you very much everyone.

kidney injury

creatinine

acute kidney injury

biomarkers

fluid management