Thank you very much for the opportunity to be here and present on this subsection of the master class regarding the use of kidney replacement therapy. Today we're going to be focused here on the use of medications and nutrition during kidney replacement therapy. I'm Erin Barreto. I'm an associate professor of medicine and pharmacy and an ICU pharmacist at Mayo Clinic in Rochester and really happy to be with you. These are my disclosures, none of which are relevant to today's presentation. When I was thinking about this talk, I was thinking about storytelling. I feel like the line is, all great stories have a beginning, middle, and end. I think with respect to continuous renal replacement therapy, we do a great job of talking about what happens at the beginning, when and in whom and how do we start CRRT. We do a good job, I guess, of talking about what happens at the end too, right? When should we transition off CRRT to another modality, bridge to recovery, or prepare for end of life? But it's the middle that I think gets a lot less attention. How do we optimize and protect the patient during the period of time that they're on kidney replacement therapy until their body is able to address their underlying condition? That's what we're going to focus on today. What's the best way to support the patient while the ship rights itself? The primary objectives for today's presentation are to discuss the PK-PD optimization of medications during CRRT, and then we're going to explain the importance of providing appropriate nutrition, vitamin, and micronutrient supplementation to these patients. I think it's easiest to think of this in the context of a patient case. Let's take a 67-year-old gentleman with hypoxemic respiratory failure, thought secondary to pneumonia and shock. He comes in with acute kidney injury and has started on CVVH, or continuous venal venous hemofiltration. What are some of the things we need to think about? With respect to CRRT, we need to consider how to avoid potential adverse effects during therapy, including avoiding hypothermia, hypophosphatemia, or other electrolyte derangements, and managing the volume effectively that the patient has on board or that we are giving iatrogenically. We also need to consider parallel interventions, which is the bulk of what we're going to talk about today, which is how we're going to approach management of the underlying condition. We think about the medications that we're going to use, oftentimes antimicrobials, during renal replacement therapy and how to make sure that they're dosed appropriately. Then we want to consider the supportive care strategies that we're offering, including nutrition therapy. First focusing on medication use, there are a number of factors that we need to consider when choosing the right drug dose for a patient on renal replacement therapy or kidney replacement therapy. These include factors related to the circuit itself, the physical chemical properties of the drug, as well as miscellaneous factors about our patient. And ultimately, these are aggregated to interpret the appropriate dosing interval for our patient that will be maximally effective and maximally safe. First looking at the physical chemical properties of the drug, there are many considerations that we want to try and pay attention to. First is the total body clearance of the medication. Renal clearance is typically only one of many possible routes that a drug could go through when eliminated from the body. If the major route of clearance for the body is non-renal, let's say for example a drug is 98% eliminated hepatically, then knocking out the kidneys with severe kidney injury is not going to have a meaningful impact on the total body clearance. So usually a good rule of thumb is something with at least 25% to 30% clearance renally will potentially be affected by our evident acute kidney injury and use of kidney replacement therapy. And fortunately, the majority of medications have this, therefore we're able to use these additional properties to interpret their dialyzability. We also want to consider the molecular weight, the protein binding, and the volume of distribution, all of which we'll talk in more detail about in just a second. And collectively, these factors are used to determine the sieving coefficient for convective clearance or the saturation coefficient for diffusive clearance modalities. The sieving coefficient and the saturation coefficient describe the ability of a solute to pass through a membrane. So when we say solute, that can relate to BUN or potassium, but it also can relate to medications. So the sieving coefficient then ranges from zero to one, one being free movement across the membrane and zero being no passage. So we use all of these factors then to approximate the sieving coefficient of a given molecule. Now to get into some of those drug properties that I mentioned earlier, the first of them is molecular weight. Molecular weight has an inverse relationship with drug clearance. The lower the molecular weight, the higher your drug clearance. So if we look at some regular small solutes, that would include urea or potassium. Most medications are going to be in this small to middle molecular weight category, somewhere between 500 and 1,500 Daltons. There are a few larger molecules, for example, monoclonals that are not going to be cleared by the dialyzer circuit, and instead their clearance would be through, for example, a different extracorporeal device like plasmapheresis. Related to molecular weight is protein binding. So high protein binding is associated with low drug clearance, and this makes sense. Both albuin and alpha-1-acetylglycol protein have high molecular weights at the 40 to 60 kilodalton size. So if they tack on to even a small molecule, it's going to reduce the ability of the molecule to get across the pores of the semipermeable membrane. So if you have a molecule that is greater than 80 to 90 percent protein bound, it's very unlikely that it is going to be cleared through the circuit. It's important, however, to acknowledge that many factors increase the free fraction of drug in critical illness, and that's because there are things like uremic toxins, altered acid-base balance, and drug interactions, as well as decreased plasma protein synthesis that can affect the proportion of protein binding that we think relative to the package insert. Still, it is more likely that a highly protein bound drug is not going to be cleared by the circuit. This is why this is often used as a poor man's sieving coefficient. If the sieving coefficient is not able to be found in the literature and it can't be calculated for the patient because you can't get samples of the ultrafiltrate or the effluent to calculate it, then you can use the protein binding to have a rough estimate. So this would be something where you would take one minus the fraction protein bound. So for example, if you had a 15 percent protein bound drug, you'd take one minus 0.15, and then the sieving coefficient would be 0.85, which would indicate, again, that the range is zero to one. So this has substantial clearance during the dialysis circuit, or in the dialysis circuit, excuse me. This study was a systematic review that aimed to determine how effectively that estimating equation predicted the true fraction unbound. So they aggregated data across studies that compared this and demonstrated that the estimated sieving coefficient and the actual fraction unbound was inconsistent across the patients included. The best performing correlation between the sieving coefficient that was estimated and the observed unbound fraction was in patients with albumin concentrations above their mean in this group. It doesn't mean that we can't still use that, but it is important to know the limitations of this estimating equation. Volume of distribution is another really important topic when it comes to drug clearance. Low volume of distribution drugs or hydrophilic medications are going to have high drug clearance relative to more tissue bound molecules. Low volume of distribution drugs will preferentially distribute intravascularly, and the renal replacement therapy circuit clears the intravascular space. There then will be modest redistribution of drug from the extravascular space into the intravascular space, which again will be cleared by the circuit. So you can see the total drug molecules that were originally present, a significant proportion of them have been removed. So again, low volume of distribution drugs have high drug clearance through the dialysis circuit. We've talked about some of the physical chemical properties of drugs that affect clearance of the circuit, but there are also things about the specific RRT circuit that we need to be considerate of. First is the modality, certainly intermittent, prolonged intermittent, or continuous have different degrees of drug elimination. Also the mechanism of solute clearance can affect drug clearance on the circuit. Convective clearance modalities are more successful at clearing minimal molecular weight molecules than diffusive clearance modalities. The dialyzer is also an important part. Older dialyzers had lower efficiency to clear small molecules and lower flux, meaning they cleared middle molecular weight molecules less. So if you extrapolate older data to today's patients where we're using high flux, high efficiency dialyzers, it would likely lead to underdosing, as the clearance today is generally higher than the clearance in the past with the contemporary dialyzers. Also you have to think about circuit interruptions or termination. So if we look at the data from the Epidemiologic DoReMi study, it suggested that the downtime during a course of renal replacement therapy was anywhere between 8% to 28% of the total treatment time, and the more downtime that you have, the less drug clearance that you're going to experience. So if you have a patient who is on and off their circuit, either due to clotting or multiple procedures or imaging, they could be at risk for occult overdosing of medications because of all of that downtime. So anybody who has greater than four to six hours of consecutive downtime, you want to make sure that you're having a conversation with your clinical pharmacist about how to reevaluate their drug doses. In addition to the modality, the dialyzer, and circuit interruptions, another thing to consider is the dialysis dose. So this is not the drug dose we're talking about, but now the dose of dialysis. When we think about dialysis dose, we think of the ultrafiltration flow rate or the dialysate flow rate. So this can be a significant factor when it comes to drug elimination. Essentially the faster that you run a dialysis circuit, the more likely you are to clear drugs. So this is a good example here on the y-axis of the left graph, you see the trough concentration for vancomycin. On the x-axis, you see the ultrafiltration flow rate. All these patients are on CBVH, and you can see an inverse relationship between the observed trough concentration and the UF rate, suggesting higher clearance of drug at higher ultrafiltration flow rates. Moreover, you can see that there's a high proportion of patients who have subtherapeutic concentrations at normal rates, even around 30 mLs per kilo per hour, which would be worrisome of course. This is further highlighted in a post-hoc analysis of the renal trial, which, as you may recall, randomized patients to high-intensity CRRT or 40 mLs per kilo per hour for their dialysis dose versus low-intensity CRRT, which was around 25 mLs per kilo per hour. And then they looked at the drug concentrations and their clearance in these two different groups. What you can see is that overall, the high-intensity group had higher drug clearance, but in most cases it wasn't clinically meaningfully different. This is reassuring that at normal hemofiltration rates, around that 25 to 30 to 35 mL per kilo per hour, we're not going to see a huge difference, but there is still significant clearance of drug overall and persistent inter- and intra-individual variability that is observed. There are a number of other considerations in addition to the physical chemical properties of the drug and the dialysis circuit that we've talked about that we need to consider. The first is residual kidney function. It is important if the patient is put on renal replacement therapy for a reason other than acute kidney injury that we consider the degree to which their existing kidney function contributes to drug clearance. For example, if you are started on dialysis for a poisoning, it's likely that you have preserved kidney function to some extent, and that in conjunction with the extracorporeal clearance and other sources of drug clearance are all very relevant, and inattention to it would lead to underdosing. We also want to be attentive to the patient's severity of illness and the therapeutic window of the medication. The mechanism of drug action is a critical piece to consider when evaluating a drug regimen. For example, with antimicrobials, we may be dealing with time-dependent versus concentration-dependent bacterial killing, and that may affect whether you choose to manipulate the dose or the drug interval. If you have therapeutic drug monitoring available, this would be a great time to use it, but even if you don't have serum drug levels, you can follow a patient's progress towards their effectiveness goals to make a determination about the appropriateness of your drug dose. For example, are their presser requirements going down? Is their white count getting better? Are their blood cultures clearing? If none of those things are true, it may not be appropriate, regardless of your understanding about what the drug dose should be, to decrease it because the patient is failing to achieve their effectiveness targets. Last, you want to re-evaluate how your performance is at your center. This is a great option for quality improvement projects. So in one center, they demonstrated that there was about 70% adherence to renal replacement therapy antibiotic dosing guidelines. They had dosing guidelines at their shop, and they demonstrated that in about a third of cases, these were not being followed for CRT, and it was especially prominent in days where a patient was changing modalities from, say, for example, CRT to IHD. In these cases, this is a great chance to have proactive conversations with the clinical pharmacist about the expected change to the modality to try and make sure that the patient's medication therapy is optimized. Wrapping this section up, then, let's return to our patient case. So again, this is a 67-year-old gentleman with hypoxemic respiratory failure and shock thought secondary to pneumonia being started on CVVH, a kind of a normal-ish prescription of around 30 mLs per kilo per hour of prescribed dose with regional citrate anticoagulation and a contemporary dialyzer. So if you were to approach the antibiotic dosing in a circumstance, we'd want to evaluate the literature and see if there was any information available to us to inform how we address and select the right dose. If insufficient data is available, which is pretty much the norm, then you want to assess the physical chemical properties of the drug, including molecular weight, protein binding, and volume of distribution. Remember, low molecular weight, low protein binding, and low volume of distribution would suggest that a drug is more likely to be dialyzable. You can estimate the sieving coefficient by one minus the fraction bound. Again, if your degree of protein binding was 30 percent, it'd be one minus 0.3 or sieving coefficient of 0.7. Then you can multiply that by the hemofiltration rate to determine what the clearance is during CRT to calculate the drug dose. You also, as I mentioned, want to consider the severity of illness, the therapeutic window of the medication, how the drug works, whether you have access to drug levels. You want to continuously evaluate your patient's response to therapy to determine if the dose needs to be manipulated beyond that which you expected. Moving on to talk about nutrition and CRT. There are a number of steps that we need to take to gauge how best to approach nutrition in critically ill patients requiring CRT. It's quite different to care for a patient with acute kidney injury requiring RRT rather than a patient with chronic kidney disease from a nutritional standpoint. In stage 3 or stage 4 chronic kidney disease, we often find that protein intake needs to be revised downward to limit the progression of kidney disease. But this is very different than the situation of acute illness or AKI requiring renal replacement therapy as we know that critical illness is associated with the catabolic state. In particular, critical illness requiring CRT results in clearance of amino acids through the extracorporeal circuit, which requires us to re-evaluate how we approach nutrition to make sure we have adequacy. The first step of any nutritional plan is to assess a patient's nutritional risk and nutritional status. You can evaluate nutritional risk using a standardized score like the Nutrix score, where a score of at least five identifies patients at high risk for nutrition. Then you can use the subjective global assessment or another tool to assess their nutritional status. Ideally, we would have enteral access for feeding. We want to make sure to understand that there's a bidirectional impact on energy requirements during renal replacement therapy. There's a loss of nutrients through the filtration and dialysis process, but also we supply substrates and other components through the circuit as well. To best approach nutrition, then we need to estimate the total calories that a patient in this scenario needs, determine their individualized macronutrient requirements with a heavy focus on protein contribution, and evaluate whether the patient has other sources of calories. Once determined, then we would select and institute their nutrition therapy program and evaluate their progress towards nutritional goals. I made specific mention of protein in the previous discussion, and this is of particular importance in patients at CRT. Current guidance is for patients at CRT to receive increased protein up to 2.5 grams per kilogram per day. These recommendations are based on at least one study that looked at caloric and metabolic balance in 50 critically ill patients. They used indirect calorimetry and ultrafiltrate nitrogen loss to determine the patient's true energy requirements. Patients were randomized to receive either two grams per kilogram per day of protein, or up to 2.5 grams per kilogram per day in the intervention group. They demonstrated that the estimated caloric requirements for patients based on indirect calorimetry was around 2,100 kilocalories per day. They showed that nitrogen balance was inversely related to energy expenditure and positively related to protein intake. As you can see here, there's a slight upward trend in the slope of this line demonstrating improved nitrogen balance with increased protein delivery. For patients specifically who received protein greater than two grams per kilogram per day, they were more likely to achieve a positive nitrogen balance, which was associated with more favorable hospital, ICU, and clinical outcomes. There are other sneaky sources of calories for patients on CRT, including drugs and diluents for medications, as well as potentially maintenance fluids that contain dextrose. A liter of D5 would contribute 170 kilocalories towards your total daily energy requirements. We also need to think of medications that are in lipid emulsions like propofol or clovidipine. Patients sedated with 35 mikes per kilo per minute of propofol can also have a significant calorie contribution at around 400 kilocalories per day depending on patient size. All of this needs to be accounted for when determining the difference that needs to be made up with nutrition therapy. Also, CRT itself can contribute calories. Some of the replacement fluids have dextrose within them, as you can see, and dextrose contributes 3.4 kilocalories per gram. Citrate itself, although we intend for it to be regional, meaning in the dialyzer circuit, some of it will get back to the patient and it too can contribute energy. This was explored in a study by Andrea Noom, published in the American Journal of Clinical Nutrition, where they looked at 10 patients and they sampled pre- and post-dialyzer to determine the degree of dextrose and citrate that was making it back to the patient to evaluate the total caloric contribution of being on circuit. What you can see in the table is that they estimate that the total net caloric gain for a patient on CRT is around 500 kilocalories per day using a CVVH circuit with contemporary parameters and settings. You can see in the graph here that the pre-filter and post-filter glucose concentrations and pre-filter and post-filter citrate concentrations were definitely higher in the post-filter setting. This is certainly something that we want to be accounting for but needs to be tailored to your circuit setup. We've talked about medication and nutrition therapy during CRT, but we can't forget vitamins and trace elements. Some of these are water-soluble molecules that are also cleared by the CRT circuit. We think about things like the B vitamins, vitamin C, zinc, copper, thiamine, selenium. Each of these may be cleared well through the CRT circuit given their molecular weight and solubility. Vitamins and trace elements are part of some very essential regulatory immunologic and antioxidant process. They function as co-factors in various enzymatic activities, including glutathione peroxidases and superoxide dismutase. If we don't pay attention to them, they're going to have some significant negative effects. Now, certainly, we know that some of these can be cleared by the extracorporeal circuit, but some of these are affected by just the strong inflammatory response in the context of critical illness. To try and differentiate that, in this one-year study in the UK at two different centers, they looked at the plasma concentrations of vitamins and trace elements in patients with acute kidney injury, a proportion of which who are on CRT and they measured these on six consecutive days. They demonstrated that all the water-soluble vitamins and trace elements were detectable in the effluent fluid. As shown on this graph, a significant proportion of patients had levels below the reference range for their vitamins and trace elements that were measured. But as you can see, if you look at AKI on CRT and AKI off CRT represented by the blue and the orange bars respectively, they're not altogether that different, which suggests that this may be more a function of critical illness rather than a function of the extracorporeal circuit. At that point, you have to wonder whether these are pathologic or whether it is just a manifestation of the ongoing event that will resolve as the patient's illness gets better. These data further look to assess the sieving coefficient and total net gain or loss of trace elements during the CRT circuit exposure. In the graphic on the left, you can see that the sieving coefficient estimated in a bovine model was relatively modest for the majority of these molecules. Remember, one would be full clearance across the dialyzer membrane and zero would be something that is not cleared at all. This is still a relatively modest degree of clearance of these trace elements. If you look at the data from a right, this was a single-center prospective crossover trial where they looked to determine trace element thiamine balances during CBV HDF, and they looked at two different replacement fluids. All of these patients got a standard multi-trace element support as well as additional thiamine supplementation. They demonstrated that there was net overall loss of copper, selenium, and thiamine and a slight net gain of zinc, which was attributed to a potential low concentration of zinc in the replacement fluids provided. Overall, what these data suggested is that even with multi-trace element supplementation and vitamin supplementation, there may be some loss, but it's going to take a significant period of CRT exposure to get there, at least more than a week typically. For the majority of our critically ill patients, they're not going to be on CRT for long enough to make these vitamin and trace element clearance issues a major concern, but in certain cases, this is going to be an issue. I bring to you this specific case reported of a 33-year-old gentleman with massive burns who ended up on CRT for almost six months beginning within the first week of admission. He received all of the regular supportive care strategies, including protocolized early trace element repletion and enteral nutrition for the first month of his admission. But as you can see on this graph in the orange diamonds and line, that by day 50, he had extremely low copper levels that persisted to be deficient despite supplementation during that first 30 days, nearly fourfold in excess of the recommended daily intake. He experienced repeated and sudden episodes of dysrhythmias, compromised hemodynamics, and you can see hypertriglyceridemia, which has been associated with copper deficiency. This is a good example of a certain circumstance where the vitamin and trace element issues become a real problem. A reasonable question then is, should we supplement or should we not, and in whom? Well, it's important to remember that fed patients, patients who get sufficient enteral or parenteral nutrition or vitamin supplementation, are going to get vitamins and trace elements from a variety of different sources. Most of these replacement thresholds are going to be more than sufficient to overcome the estimated 24-hour balances that have been shown in the literature in patients on CRRT. Also, we must recognize that low levels may not require supplementation. These may be just a function of critical illness, and actually supplementing them may have no meaningful impact on the patient's clinical outcomes. If you have a patient who's on continuous renal replacement therapy for more than one to two weeks, this may be a time where you want to start thinking about checking levels of these vitamins and minerals, and then making an individualized decision about whether you want to supplement. Recognizing that high doses of vitamin C in particular may have some association with acute kidney injury, and so you want to be conscientious about doing that. Bringing it back to the patient that we've been following then, if we look at how we would treat this particular case, we would appraise him as high nutritional risk given his severity of illness and past medical history. It would be appropriate to start enteral feeding within the first 48 hours if he's not able to feed volitionally, which we're expecting to be the case. If you estimate his goals using either a estimating equation or if you have access to indirect calorimetry, we could estimate that maybe he needs around 1,500 kilocalories per day. If he's on CRRT, we would project that a protein intake of around 1.8-2.5 grams per kilogram would be appropriate, with 30 percent of calories as fat and the remainder as dextrose-containing fluids. We'd also want to look at calories from other sources, including fluids and meds and CRRT, and we might ballpark that he's getting 500 kilocalories per day from those sources, and around 250 from his propofol sedation, which would leave us with a deficit of around 750 calories that we need to make up and some amount of protein that we need to increase. You could use this to then prescribe a specific enteral nutrition program that would be tailored to his overall requirements. Based on the epidemiologic data from the best kidney group, the estimated duration of CRRT for the majority of patients is around six days. Assuming the patient gets the enteral feeding that we intend within around the first 48 hours, he will receive the recommended daily intake of vitamins and trace elements, which would be sufficient to overcome any net loss through the circuit. However, if CRRT were to continue for at least two weeks or more, then we might consider checking, trending, or supplementing the trace elements. Now, there are obviously certain subpopulations that we want to be attentive to. As I showed, burns is a particularly risky group. We also want to think about patients who have major abdominal surgery where the total caloric requirements as well as vitamins and trace minerals may be uniquely affected. To wrap up then our discussion on medication and nutrition therapy during CRRT, it's important to remember that CRRT decisions are not just about starting and stopping. That's certainly where a lot of the literature is, but how we maximize effectiveness and safety during therapy is just as important. When thinking about things that are cleared by a dialysis circuit, low molecular weight molecules with low protein binding and a low volume distribution are much more likely to be eliminated by the circuit. But you also want to think about the CRRT intensity, the patient severity of illness, and other factors to determine the appropriate dose and interval for a given medication. Nutrition therapy during CRRT needs to be individualized, and you need to acknowledge that there's a bidirectional impact of the circuit on energy requirements. The circuit itself can provide some molecules to the patient that contribute energy. We talked about dextrose and regional citrate as potentially issues, but also some other water-soluble molecules, including vitamins and trace minerals, as well as amino acids can be cleared. When you're calculating the patient's energy requirements during CRRT, factor in those other sources of calories and consider a protein load of around 1.8-2.5 grams per kilogram per day. Finally, remember that generally, the recommended daily intake of vitamins and trace elements will be achieved through regular nutrition therapy. However, if you have a patient who is on prolonged CRRT, then you may want to screen for deficiencies. But know that these are critically ill patients and deficiencies are quite common and may not be uniquely affected by the extracorporeal circuit, nor may supplementation be beneficial. It's been my privilege to be with you today and share this material. I hope it was helpful. But please do reach out to me if you have any questions. I'm available by e-mail or on my Twitter here. I'd be happy to talk through it in more detail. Thank you.

medications

nutrition

kidney replacement therapy

optimizing patient

drug dose determination

protein intake

supplementation