Frank Maddux: Fluid management is a critical component of dialysis and has come a long way. Fresenius Medical Care's Kidney Disease Think Tank, the Renal Research Institute, developed its own ultrafiltration controller. This innovation continuously compares the patient's relative blood volume profile to the target curve and makes ultrafiltration rate adjustments. Dr. Lemuel Rivera Fuentes is Supervisor of Clinical research at RRI. He joins us to discuss this innovation and its designation as a U.S. Food and Drug Administration 21st Century Breakthrough Device.
Frank Maddux: Lem, thanks for joining me. Talk to us just a little bit about the challenges of fluid management and volume control during dialysis, and how this is such an important topic.
Lemuel Rivera Fuentes: As we know in healthy patients, our kidneys function all day long, twenty-four-hours, seven. And this makes it very easy to manage fluid, sodium and all the other things that regulate volume in patients. When we lose kidney function, then we lose this regulation. And whatever the patient has as an intake of fluid, it's going to stay there unless we take it out. And the issue is that, of course, we have patients that are coming to our clinics three times a week for a very specific amount of time. And we're trying to get as much out of the treatment as we can for our patients, which means to put these patients through a very stressful four hours, where we're trying not just to regulate all other levels as a base, etcetera, but we're also trying to take as much fluid as we can in order to get them to euvolemia. So, we're packing a lot in four hours' treatments. And that of course has consequences for the patient, because these patients potentially are very sick patients with traditional risk factors for end-stage renal disease, which among them the cardiovascular risk factors are very important. Especially as they increase in age. And cardiovascular function and fluid management go hand-in-hand. They feed off each other and they really have a very dependent behavior in our patients and in the treatment that we give to our patients on a regular basis. SO, yeah, this creates a very intricate balance between what we can do for our patients, and what they actually can go through in a very short amount of time. It makes it very complicated to get through exact volume management that we need from them.
Frank Maddux: Cardiovascular health, as you know, is an important pillar in our clinical and quality agenda. Let's talk a little bit about patient outcomes and the impact of intravascular volume or relative blood volume changes during a dialysis treatment. Why are those outcomes impacted by that?
Lemuel Rivera Fuentes: Well, as we know chronic volume overload puts a huge amount of strain in our cardiovascular system, in cardiac function and in vascular elements. So, the way that we need to target the treatment in these patients is particularly challenging. We need a better way to define what is volume in these patients. And the best way to manage this volume chronically through their treatments in dialysis. relative blood volume has come up as a tool that has helped us guide fluid management and improve cardiovascular stability by doing online measurements of what is potentially our extravascular cellular volume in real time during dialysis treatment. And estimating our vascular refilling capacity, which is kind of like a secondary measurement of our fluid overload patients. It's a very simple way to try to understand what's the dynamic between our fluid overload in patients and their cardiovascular stability during treatment. We're trying to use it, on a daily basis to increase potential outcomes in the future for these patients regarding, fluid overload and volume.
Frank Maddux: And certainly interdialytic hypotension is a substantial risk for patients as well. Tell me a little bit about what we know about the algorithm related to the adaptive UF control. The controller itself is essentially based on this algorithm. Can you give me a sense of how you guys developed this algorithm?
Lemuel Rivera Fuentes: It all came out of these observational studies that we did with around 800 patients where we were trying to figure out what was the difference between patients that actually had a better survival of 30 months after our study. And patients that, unfortunately, were not as fortunate. And what we realized is that the way that their RBV curve behaved during the treatment was completely different. Patients that had a better survival at six months or thirty months had a steeper slope in their RBV curve during the treatment that was achieving certain targets. Every 30 minutes, every hour of the treatment, up to the point that at the end of the treatment they had achieved a specific reduction or at least a specific range of reduction in RBV curve. Which means that these patients potentially have a little better fluid management during these treatments. So, what we did was try to figure out a way to create a controller that would steer patients RBV curves into these predefined target ranges that were associated with a better survival and a better outcome. And this is something that, of course, in clinical practice was a little bit difficult, but we managed to create this algorithm that continuously every ten minutes gives us different UFR rates that which goal is to steer the patients into these predefined benefit ranges of RBV.
Frank Maddux: So it's quite clear that our population of patients benefit from these devices that are on the machine that can help create enhanced safety and actually full delivery of what the desired treatment is. And trying to find that optimal level of blood volume and yet not cause hypotension or difficulty in intravascular volume is really one of the tricks that I think a more advanced machinery actually will help provide a safer treatment. This Breakthrough Device designation. What is that? Can you just tell our audience a little bit about the details of what the Breakthrough designation provides?
Lemuel Rivera Fuentes: The Breakthrough Device designation for certain medical devices that provide a more effective treatment or diagnoses of life-threatening or irreversible debilitating diseases or conditions. The goal of this program is to create a clear pathway for these devices to enter into the market. And this means a closer interaction with the FDA in all premarketing activities and a clearer pathway for review of the device. This, of course, is meant to help getting these devices into patient care and healthcare providers as soon as possible, because they are deemed as breakthrough discoveries that will help eventually patients in real life. And we were able to obtain this designation early-on with our algorithm.
Frank Maddux: Fluid status for clinicians has always been a bit of the art of nephrology and doing that assessment of a patient. How does this as a diagnostic tool actually add to the armamentarium of a nephrologist trying to prescribe, not only an effective therapy, but a very safe therapy. Are there other things that you all are working on?
Lemuel Rivera Fuentes: This is always in the discussion around, of course, volume management, how best to define what is euvolemic or hypervolemic and how best to describe it in the patient. Every patient is so different. For the longest time we were just taking into consideration clinical assessment of the patient, which of course, we now know that it's not accurate enough for the treatment of our patients. And as technology has evolved, we have been able to come up with different diagnoses methods, or assistance like RBV, for example, monitoring during real life treatments, that have helped shaped our understanding of volume management. We're currently working with VCN we're currently working with the RBV curve, we're working with different other interdialytic algorithms and also predictive models that eventually will shape a little better our understanding of fluid management and then of the consequences of this fluid management in patients during treatment. It's a little bit early for me to say specifically what we are working on but rest assured that we're trying to implement all the armament that we have. Not just in terms of clinical knowledge, but also in terms of mathematical power and statistical analysis to try to bring everything together. Because at the end of the day, the best outcome of our patients is going to be our proper, or our accurate use of all the devices or all the tools that we have at our disposal. It's not that one tool is better than the other one. It's more that we need to take into consideration all of these tools and use them together in each of our patients.
Frank Maddux: It's quite clear to me as a clinician that absolute blood volume and what you might get from the body composition monitor or some equivalent device and relative blood volume that gives you a sense of how things are changing in the intervascular space are both needed to effectively try to help squire an individual patient through a treatment safely. Renal Research Institute has used virtual clinical trials and other in silico computerized simulations quite effectively at trying to help us look at medications and other things. Has it helped at all in testing this UF controller?
Lemuel Rivera Fuentes: In early stages of the development of this algorithm in silico testing was required, because of course, it's faster to test these kind of theories when you do it virtually. It helps tuning the controller area, but it requires relative simulation, so of RBV slopes. So, what we did was we took a physiology based model of plasma refilling and we adapted to 13 individual patients following the methods that are described elsewhere. And what we do is we run simulations of different treatments in these patients and it really helps us fine-tune the gains and fine-tune the specifics of the algorithm. We've done these in so many other areas within the treatment of our patients, for example, anemia. In this case, volume management. And it really is a testament of how far we have advanced in mathematics and in programming and in data analysis. Because now what we can do is basically create a virtual representation of patients and then test all these theories or all these models or all these algorithms once again and once again with different physiological values or different physiological conditions before actually getting these algorithms to patients in real life. And once these algorithms or these tools arrive to the patients, we can be certain that a good amount of security measurements have been taken into consideration for the creation of these tools. So, yeah, it's been intrinsic for the development of such cutting-edge technology.
Frank Maddux: You were one of our authors in this year's Annual Medical Report on this particular controller and device. Describe to me a little bit this how you involve nurses in the study of trying to figure out where the right check and balances in assessing a patient and turning over some of the responsibility to this automated algorithm.
Lemuel Rivera Fuentes: Our first concern always is the patient's safety. And since this was a pilot study, which is a first approach to how this would look like with a real-life treatment, it was very important for us to have a safety mechanism before actual implementing a fully automated version of our algorithm. What we came up with was to have this nurse in the loop as we call it. Which means that our current algorithm, it gives UFR changes every ten minutes, or suggestions of changes every ten minutes. And what we did was that every ten minutes we were having these ultrafiltration rate changes, but before they were implemented into the actual patient we have a nurse in the loop and determining the area between what our simulation or what our computer programming was telling us, and what was actually happening in real life in the clinical setting of the patient. So, it was the job of the nurse in the loop to figure out if the recommendation that we were providing was actually safe for our patient and if it was implementable for the patient in a real-time scenario. So, they were accepting when the clinical experience of the nurse in the loop dictated, so they were not accepting the recommendations. It was impressive for us to know that almost 90 percent of the recommendations that we gave during our-- after the course of our study, were actually followed through the nurse in the loop. And yeah, it was a very good way to start testing this algorithm with the safety regulations that we definitely needed to have in order to protect our patients. Yeah, it was an amazing experience and a very good first dip into what this potentially might look like eventually once we have fine-tuned everything that we need to fine-tune in this technology.
Frank Maddux: So, let me put you on the spot for a minute. What do you see as the future for the UF controller?
Lemuel Rivera Fuentes: I think the future of the UF controller will need several other variables to be involved in the decision-making process. In this case, for example, we know that the balance between how much fluid I am taking out of the patient and how stable my cardiovascular system is, it's so difficult to achieve. So, we would need for sure eventually to think about variables that define these kind of events. In this case, for example, I think that adding BP measurements, or I think the prediction of how those BP measurements might translate into IVH potentially in patients is something that we currently are looking into and we're going to, for sure, add. And also, what we did with this first pilot study was that we selected an ultrafiltration goal for the patient in the treatment, but we also selected a range, an upper level or a maximum ultrafiltration goal, and a minimal ultrafiltration goal, in order to steer the patient into this particular range. All these kind of questions of individualizing therapy are going to definitely play a very important role in the future of what we're going to come next with ultrafiltration controller for sure. Trying to fine-tune it, individualize it even more and take into consideration other variables that for sure have a big say in how our patient behaves during volume management in hemodialysis or chronic hemodialysis.
Frank Maddux: One last question before we finish and it's on cardiovascular health for our population. Give me your thoughts about whether you think we should be monitoring the electrocardiogram and the rhythms of our patients during treatment.
Lemuel Rivera Fuentes: I think there's two ways to look at it. It's definitely the more data we have about patients in real-life, of course, the better, because we tailor our treatments and we tailor our approach to patients using these real-life data. Saying that, though, it is difficult sometimes to know what to do with all this data. Especially because at the end of the day we need very practical decisions that can be followed up and can be applied in populations, in big population basis. Something that is not scalable up shouldn't be something that we prioritize. In this case, also, for example, taking EKGs of patients in real life during the treatment, if we don't know what we are going to use that information for, we potentially should-- take it with a little bit of grain of salt. And this means that if my patient is doing an arrhythmia during the treatment, it's very good to know if the patient is doing so. But also, if the arrhythmia is not symptomatic and is going to hinder my volume management, what is the correct balance between having the information and using the information or having the information and not using the information? It's a very tricky question. The more information we have, the more we like, and the more interesting decisions and interesting propositions we may come up with, but we need to be very careful into defining exactly what we're looking out of this information. And if there is no clear purpose or clear plan of action with that, I would be reluctant to scale it up to all patients at once.
Frank Maddux: Thank you. I'm here today, and our guest has been Dr. Lemuel Rivera Fuentes from the Renal Research Institute, talking about volume management and ultra-filtration control. Lem, thanks so much for joining me today.
Lemuel Rivera Fuentes: Thank you so much.