We had the privilege of sitting down with Prof. Sean Bagshaw from Edmonton, Canada to discuss Renal Replacement Therapy in the ICU.
Dr. Bagshaw is Professor and Chair, Department of Critical Care Medicine at the University of Alberta in Edmonton, Alberta, Canada. He completed medical school and residency training at the University of Calgary (Doctor of Medicine – 2000; Internal Medicine – 2003; Critical Care Medicine – 2005; Master of Science in Epidemiology – 2005) prior to completing a Critical Care Nephrology fellowship in the Department of Intensive Care Medicine, at the Austin Hospital in Melbourne, Australia. Dr. Bagshaw works as a full-time staff intensivist in two of the busiest and highest acuity intensive care units in Alberta – the General Systems ICU at the University of Alberta Hospital (2007 – current) and the Cardiovascular Surgical ICU at the Mazankowski Alberta Health Institute (2011 – current).
The following blog is based on that conversation.
Initiation of Renal Replacement Therapy
Prior to STARRT AKI, the evidence was captured in a few landmark trials, which were mostly single centre and not multinational. In the ELAIN trial, mortality benefit was demonstrated for early initiation of renal replacement therapy in critically ill patients with AKI (2). This trial had its problems though, most notably external validity due to its single centre nature as well as a fragility index of 3, which meant that with a shift of 3 patients, the outcome would change to non significance. In the French multicentre IDEAL ICU trial, no benefit was shown with early initiation of renal replacement therapy for septic shock patients (3). In fact, the trial was stopped early due to futility. Another French trial, the AKIKI trial showed no mortality benefit at 60 days with early renal replacement therapy in critically ill patients (4), which led to STARRT AKI, which was the largest trial to consider timing of renal replacement therapy in critically ill patients.
STARRT AKI was a multinational collaborative effort recruiting approximately 3 000 patients across 15 countries in 168 centers using a strategy of accelerated initiation of renal replacement therapy, versus a standard approach amongst critically ill patients with acute kidney injury (1)
Firstly, no difference was found in 90 day all cause mortality, which was the primary end point across several sensitivity analyses. The second major finding was that greater renal replacement therapy dependence was found at 90 days amongst patients who had been allocated to the accelerated strategy. This was a robust finding in a series of sensitivity analyses. The third finding was that there was a greater occurrence of adverse events amongst those allocated to the accelerated strategy, probably driven by episodes of hemodynamic instability. The final finding was that 38% of all patients allocated to the standard strategy did not go on to receive renal replacement therapy. This to some extent may have some implications for resources and/or dialysis provision in critical care settings (1)
It is important to note that patients with classic indications for urgent renal replacement therapy, were largely excluded from these trials. The question then arises whether one, or a combination of these classic indications, would be sufficient to initiate renal replacement therapy since some of the indications may be medically managed. From the AKIKI trial, it was clear that patients with a delayed strategy were more likely to receive diuretics to manage fluid, insulin to manage hyperkalemia and bicarbonate to manage acidosis (4). Did this diminish the risk of renal replacement therapy? Possibly. What is not known from the literature is whether these interventions pose a hazard to patients.
In the large multicenter French trial, BICAR ICU, the question of giving bicarbonate to patients with acidosis in the ICU was looked at. In a subgroup of patients in this trial, it was suggested that there was a decrease in the utilization of renal replacement therapy (5). Should we medically intervene and watch and wait for a period? Maybe, however there may be a point of diminishing returns. STARRT AKI and other clinical trials suggest that a delayed or standard strategy may be acceptable in the short term. What is not clear from the literature, is whether our interventions to avoid dialysis may be harmful. The question of initiation is not completely settled.
Who gets to make these nuanced decisions for critically ill patients? Intensivists are experts in critical illness, multi-organ failure and multi-organ support management. Nephrologists are experts not only on the kidney and all aspects of kidney disease, but clearly on dialysis – the physiology and biomechanical aspects thereof. There is expertise on both sides, and both specialties need to be engaged in the care of these complex patients.
Once patients improve enough to be liberated from life support measures, patients with acute kidney injury supported with extracorporeal support in the form of dialysis, are either recovered or not when they are handed to nephrology services. Even if they’re not receiving dialysis anymore, we know that there are implications downstream after having had a severe episode of acute kidney injury on long-term kidney health and survival. ICU may not be doing a great job at transitioning these patients to the nephrology services.
Which is the ideal Modality of Renal Replacement Therapy in the ICU?
When it comes to renal replacement therapy in the critical care setting, there remains controversy around the ideal modality to support patients with kidney injury. Continuous therapies were developed to replace intermittent therapies in patients with intractable hemodynamic instability. In general, ICU should be able to offer all modalities of support that suit the specific patient’s phase of illness. There are clear scenarios where continuous renal replacement therapy is likely the best initial therapy, on the other hand, there are scenarios where there is no doubt that an intermittent modality with high efficiency clearance is the better. Scenarios such as overdoses or poisonings with dialyzable toxins, intermittent therapy should be the choice!
Evidence has not shown survival benefit between intermittent or continuous modalities in patients with haemodynamic instability or multiorgan dysfunction. Even though clinical trials haven’t shown a clear difference in survival, there may be some benefit for long-term kidney recovery and independence from dialysis amongst those patients who get at least continuous therapies in the ICU as their initial modality. If you think about the trajectory and phases of illness that a patient endures, realistically speaking you’re not going to awaken and mobilize a patient around the unit in the first 24/48 hours. So, the issues around the drawbacks of CRRT, which have largely been centered around not being able to mobilize patients, is probably a moot point.
Once patients are weaning life support and become a little more stable, but still in need of renal support, a low efficiency, prolonged intermittent therapy may be provided that allows delivery of renal replacement therapy with our goals in mind. Patients who don’t recover renally, even as their organ dysfunction and dependence on other life support modalities diminish, may still be dialysis-dependent and will transition to intermittent therapy.
What is the ideal “dose” of Renal Replacement Therapy in the ICU?
Once the decision is made to initiate dialysis, “aggressive” prescriptions have no benefit over conservative prescriptions. Two large clinical trials, RENAL, ATN (6,7) and a series of other small trials have shown no incremental benefit in terms of survival with a higher intensity prescription of continuous renal replacement therapy in the ICU. We have gravitated to a default in CRRT around 25cc/kg/hr and intermittent therapies of 1.2 spKt/V every other day.
It is important to point out that those trials focused on solutes only and not ultrafiltration. A higher intensity has no benefit. At minimum, we have a floor at which to prescribe renal replacement therapy. This does not mean that clinicians are not mindful of the changing physiology and metabolic profile of patients. Generally, we have strong evidence on what the baseline prescription should be.
Insofar as hemofiltration, hemodiafiltration or haemodialysis is concerned, there have been some small trials. None of them showed a survival advantage among these forms of dialysis. Patients who are allocated to continuous hemodiafiltration however, seem to have longer lifespans of their circuits and filter sets. Just hemofiltration with its dependence on gradients across the membrane and associated “caking” of proteins in the filter, results in earlier circuit loss. To some extent the default is hemodiafiltration to prolong filter lifespans.
How should we manage ultrafiltration?
How do we manage ultrafiltration? How do we target fluid removal? How do we know how fast or how slow to remove fluid? These are questions that are not deeply seated in evidence. Randomized control trials are being developed and will probably be deployed in the next few years. Some observational post-hoc data from trials such as RENAL (6) have suggested that earlier initiation of renal replacement therapy can better manage fluid balance and result in lower accumulation of fluid over time. It does not necessarily translate into improved survival as shown in STARRT AKI (1).
In the RENAL trial, it was shown that really low and really high rates of ultrafiltration may exert some hazard on patients. In a secondary analysis of RENAL (6), a net ultra filtration rate exceeding 1.75cc/kg/hr vs. 1.01cc/kg/hr was associated with lower survival. Ultrafiltration and fluid management are critical pieces missing high-level evidence right now.
Is there a preference for anticoagulation?
Anticoagulation is needed to preserve the extracorporeal circuit. Patients who have frequent disruptions and loss of their circuits are more likely to lose blood and therefore more likely to require blood transfusion. If you don’t use anticoagulation, you’re not going to have as long a circuit lifespan, with the average being somewhere in the range of 18 to 20 hours with no anticoagulation. With Heparin it is a bit longer and with Citrate, it is even better.
Citrate is a little more complex, but protocols are in place so that they may be applied safely as has been shown in observational studies and clinical trials. The risks of Heparin, being systemic, are that of bleeding and Heparin-induced thrombocytopaenia, albeit a small risk. If patients have an indication for systemic Heparin, then it is usually a good choice.
There was some evidence in the COVID 19 pandemic where both modalities of anticoagulation were being used due concerns about circuit loss in patients secondary to a prothrombotic effect related to COVID 19. This was either due to an intrinsic coagulation disorder related to COVID 19 infection or other factors. Using both modalities, however, didn’t seem to be beneficial.
The catheter is another critical element to the whole apparatus in providing high quality and high efficiency renal support. If you’re losing your circuit, have a look at your catheter to make sure your catheter is not a problem.
How do we best liberate patients from Renal Replacement Therapy?
There has been a paucity of literature on liberation from renal replacement therapy in the critically ill. The only evidence seems to be from a small series of observational studies. It is a critical knowledge gap in our provision of renal support in the ICU. When to stop and how to get the patient off dialysis is important. Patients who remain dialysis dependent have a spectrum of complications downstream with associated worse clinical outcomes. We should be more focused on the timely liberation of renal support just as is the case for mechanical ventilation.
In general, one must consider the phase of critical illness, the resolution of organ injury and the removal of the stimulus that incited the acute kidney injury. There should be some evidence of intrinsic kidney recovery. The only barometer we use for the most part at the bedside is urine output. Urine output may be a sign that the kidneys are recovering and that the demand supply mismatch that has occurred, has improved. Once patients start to urinate freely, then one may be able to titrate or tailor back on there renal replacement therapy.
Does Furosemide play a role?
When the question arises on whether Furosemide would liberate the patient from therapy or prevent Renal Replacement Therapy, the short answer is no. Lasix will not improve kidney function and will not help the kidneys clear more solutes. It may allow volume management, but that’s about all. After elective discontinuation of dialysis, titrating Furosemide is not going to equate to improved probability of recovery.
What are some future directions when it comes to Renal Replacement Therapy?
The future of Critical Care Nephrology is exciting as we further address knowledge gaps. Some greater precision work will be done in the future on who to select and how to target renal replacement therapy. Further work is also expected on fluid management and ultrafiltration, such that we are not contributing to progressive fluid accumulation, nor making patients hypovolemic and contributing to the propagation of organ injury. Furthermore, there is work being done on how to decide when to provide patients with a trial of liberation from renal replacement therapy. Although we focused earlier on urine output, there are several other markers being investigated to provide some clinical decision support on the probability of successful liberation from dialysis therapy.
Finally, the legacy of acute kidney injury in the ICU needs to be addressed. Intensivists and the critical care nephrology community have not done a very good job at following up on what happens to patients who had acute kidney injury in the course of their critical illness. What the long-term kidney health effects are after a severe episode of AKI and provision of dialysis in the ICU is largely unknown.
Bibliography:
1. Timing of Initiation of Renal-Replacement Therapy in Acute Kidney Injury. New England Journal of Medicine. 2020 Jul 16;383(3):240–51.
2. Zarbock A, Kellum JA, Schmidt C, van Aken H, Wempe C, Pavenstädt H, et al. Effect of Early vs Delayed Initiation of Renal Replacement Therapy on Mortality in Critically Ill Patients With Acute Kidney Injury. JAMA. 2016 May 24;315(20):2190.
3. Barbar SD, Clere-Jehl R, Bourredjem A, Hernu R, Montini F, Bruyère R, et al. Timing of Renal-Replacement Therapy in Patients with Acute Kidney Injury and Sepsis. New England Journal of Medicine. 2018 Oct 11;379(15):1431–42.
4. Gaudry S, Hajage D, Schortgen F, Martin-Lefevre L, Pons B, Boulet E, et al. Initiation Strategies for Renal-Replacement Therapy in the Intensive Care Unit. New England Journal of Medicine. 2016 Jul 14;375(2):122–33.
5. Jaber S, Paugam C, Futier E, Lefrant JY, Lasocki S, Lescot T, et al. Sodium bicarbonate therapy for patients with severe metabolic acidaemia in the intensive care unit (BICAR-ICU): a multicentre, open-label, randomised controlled, phase 3 trial. The Lancet. 2018 Jul;392(10141):31–40.
6. Intensity of Continuous Renal-Replacement Therapy in Critically Ill Patients. New England Journal of Medicine. 2009 Oct 22;361(17):1627–38.
7. Intensity of Renal Support in Critically Ill Patients with Acute Kidney Injury. New England Journal of Medicine. 2008 Jul 3;359(1):7–20.