Recently two trials examining the efficacy of the early utilization of renal replacement therapy (RRT) in critically ill adults with acute kidney injury (AKI) in the ICU were published. Despite examining very similar patient populations, undergoing very similar modes of therapy, these trials produced drastically divergent results. The following is a brief exploration of each respective trial and the possible reasons behind their incongruity.
The first of these trials was published online on May 15th in the NEJM (1). Entitled the Artificial Kidney Initiation in Kidney Injury (AKIKI) trial, Gaudry et al randomized 620 critically ill adults with stage 3 AKI, as per the KDIGO criteria, to either immediate RRT, or the addition of renal support only when medically necessary (severe hyperkalemia, metabolic acidosis, pulmonary edema, or a blood urea nitrogen level greater than 112 mg/dL, or oliguria for more than 72 hours). The type of renal replacement therapy utilized was left to the discretion of the treating physician.
The authors found no difference in their primary outcome, 60-day mortality, between the patients randomized to immediate or delayed therapy (48.5% vs 49.7% p-value of 0.79). Additionally, no difference was observed in the rate of death at 30-days, ICU length-of-stay, long-term dialysis dependence, ventilator-free days, and vasopressor-free days between the early or late groups. The authors did report a difference in the number of patients who eventually required RRT during their ICU stay. 98% of the patients in the immediate therapy group underwent RRT during their ICU stay, while only 51% of the patients in the delayed strategy ended up requiring renal therapy. The majority of these patients (68%) were dialyzed because of either oliguria for > 72 hours, or a BUN greater than 112 mg/dL. Additionally, patients randomized to the delayed group experienced adequate urine production (defined as 1000 mL of urine in a 24-hour period without the help of diuretics) earlier than those randomized to immediate renal support. Not surprisingly, the patients randomized to immediate dialysis experienced a higher rate of catheter-related bloodstream infections (10% vs 5%).
And while this was a large high quality RCT, these results are in direct contrast to the second study published one week later by Zarbock et al in JAMA (2). The authors of the early versus late initiation of renal replacement therapy in critically ill patients with acute kidney injury (ELAIN-Trial) randomized 231 critically ill adults with stage 2 AKI to either immediate dialysis or renal replacement therapy only after meeting stage 3 KDIGO criteria. Once the decision to institute RRT was made, all patients received continuous venovenous hemodiafiltration and were only transferred to intermittent or sustained low- efficiency dialysis if renal replacement was required for more than 7 days.
Contrary to the AKIKI cohort, the ELAIN trial found a significant improvement in their primary endpoint, 90-day mortality, in patients who were randomized to early RRT. The 90-day mortality in patients who received early RRT was 39.3%, compared to 54.7% in patients who were randomized to the late group. 28-day mortality was 30.4% and 40.3% respectively. Patients randomized to early dialysis experienced shorter durations of mechanical ventilation (median of 125.5 hours vs 181.0 hours). Furthermore, there was a 7.8% absolute increase in the number of patients who still required RRT at 60-days (15.9% vs 23.7%), after randomization in the patients who received delayed initiation of renal support, though recovery of renal function at 90-days was statistically equivalent (88.2% vs 85.2%).
So how do we reconcile these two trials with drastically different results? Was it a difference in timing of therapy? Patient in the AKIKI cohort were not randomized until they were in a more severe state of kidney injury. Furthermore, the threshold for initiation of RRT in the delayed group was far more conservative than the patients in the ELAIN cohort. In fact, only 51% of the patients in the delayed arm of the AKIKI trial ended up receiving extracorporeal renal support where 91.6% of the patients in the delayed arm of the ELAIN trial underwent RRT. One could argue that it was the earlier initiation of RRT in the ELAIN cohort that led to the clinical improvement. But the fact that 50% of the patients in the delayed arm of the AKIKI cohort never received RRT without any obvious clinical consequences makes this possibility less likely.
Was it the type of RRT that was initiated? Continuous renal replacement therapy (CRRT) was used exclusively in the ELAIN cohort, while only 45% of the patients in the AKIKI trial received CRRT as their initial form of RRT. Though this could explain the difference in apparent efficacy between the two trials, large high quality RCT data has failed to show a benefit for CRRT when compared to intermittent dialysis (3).
Was it the type of patients enrolled in each respective study? Approximately 80% of the patients enrolled in the AKIKI cohort were admitted to the ICU because of a medical malady (the vast majority of these were sepsis). In contrast, the majority of patients enrolled in the ELAIN cohort were admitted to the ICU following a surgical intervention. One could argue that AKI secondary to a surgical insult is a far different pathologic process than AKI caused by sepsis.
While each of these differences seem plausible, there is far too much statistical noise to draw decisive physiological conclusions based only on these two cohorts. Certainly from a Bayesian perspective, the small sample size and single center nature of the ELAIN dataset make it far more vulnerable to random errors in statistical chance. Especially given that the prior evidence examining this question has primarily failed to demonstrate efficacy of early initiation of RRT (4,5,6). More importantly, the benefits of early RRT observed in the ELAIN cohort are far too pronounced for the clinical context in which they were produced. The median time difference to initiation of dialysis between the immediate and delayed groups was 25.5 hours. This small clinical difference leading to a 15.4% absolute decrease in 90-day mortality is implausible, especially in light of the uniformly negative RCT data published prior to this trial. A far more likely explanation for the large benefit observed with early initiation of RRT in this cohort is random errors in sampling.
This of course speaks to the greater question, how best to manage AKI? Previous trials examining therapies intended to directly act on the kidney, like forced diuresis or RRT, have failed to demonstrate any renal sparring effects (7,8,9,10,11,12,13). Maybe it is time to accept AKI as a non-modifiable marker of poor outcomes, a surrogate for the severity of illness. Attempts to directly manipulate this solitary system does not improve clinically important outcomes. Rather than focusing on the kidney in isolation, we should redirect our resuscitative efforts on the underlying disease responsible for the renal injury.
Early initiation of RRT in the AKIKI cohort led to a faster and more consistent correction of the laboratory defined kidney injury. But this correction did not lead to improvements in clinical measures of renal function or overall survival. In fact, patients randomized to the delayed group experienced an earlier return of spontaneous clearance of creatinine, despite the sustained elevation in laboratory values indicative of kidney dysfunction. In the resuscitation of the critically ill we preferentially lean towards normalcy or uboxia. Forever attempting to achieve a logical fallacy, that if we drive physiologic variables towards their assumed value in the healthy host, we will avoid the poor outcomes associated with these physiological perturbations. Correcting deviations in surrogate endpoints does not directly translate into improvements in clinically important outcomes.
*CC Nerd is EM Nerd's slightly less nihilistic alter ego. Topics that have a more Critical Care bend will go under the tittle of CC Nerd. Hopefully these new adventures will be a welcome addition to the EM Nerd Chronicles.
Sources Cited:
- Gaudry S,Hajage D, Schortgen F, et al. Initiation strategies for renal-replacement therapy in the intensive care unit. N Engl J Med.
- Zarbock A, Kellum JA, Schmidt C, et al. Effect of Early vs Delayed Initiation of Renal Replacement Therapy on Mortality in Critically Ill Patients With Acute Kidney Injury: The ELAIN Randomized Clinical Trial. JAMA. 2016;315(20):2190-9.
- Vinsonneau C,Camus C, Combes A, et al. Continuous venovenous haemodiafiltration versus intermittent haemodialysis for acute renal failure in patients with multiple-organ dysfunction syndrome: a multicentre randomised trial. Lancet 2006;368:379-385
- Wald R, Adhikari NK, Smith OM, et al; Canadian Critical Care Trials Group. Comparison of standard and accelerated initiation of renal replacement therapy in acute kidney injury. Kidney Int. 2015;88(4):897-904.
- Wald R, Adhikari NK, Smith OM, et al; Canadian Critical Care Trials Group. Comparison of standard and accelerated initiation of renal replacement therapy in acute kidney injury. Kidney Int. 2015;88(4):897-904.
- Jamale TE,Hase NK, Kulkarni M, et al. Earlier-start versus usual-start dialysis in patients with community-acquired acute kidney injury: a randomized controlled trial. Am J Kidney Dis2013;62:1116-1121.
- Cantarovich, B. Rangoonwala, H. Lorenz, M. Verho, V.L.M. Esnault, High-Dose Furosemide in Acute Renal Failure Study Group. High-dose furosemide for established ARF: A prospective, randomized, double-blind, placebo-controlled, multicenter trial. Am J Kidney Dis, 44 (2004), pp. 402–409.
- Cantarovich F, Fernandez JC, Locatelli A, Perez Loredo J: Frusemide in high doses in the treatment of acute renal failure. Postgrad Med J 147:13-17, 1971.
- Cantarovich F, Galli C, Benedetti L, et al: High dose frusemide in established acute renal failure. Br Med J 4:449-450, 1973.
- Brown CB, Ogg CS, Cameron JS, Bewick M: High dose frusemide in acute reversible intrinsic renal failure. A preliminary communication. Scott Med J 19:35-39, 1974.
- Kleinknecht D, Ganeval D, Gonzalez-Duque LA, Fermanian J: Furosemide in acute oliguric renal failure. A controlled trial. Nephron 17:51-58, 1976.
- Brown CB, Ogg CS, Cameron JS: High dose frusemide in acute renal failure: A controlled trial. Clin Nephrol 15:90-96, 1981.
- Shilliday IR, Quinn KJ, Allison ME: Loop diuretics in the management of acute renal failure: A prospective, double-blind, placebo-controlled, randomized study. Nephrol Dial Transplant 12:2592-2596, 1997.
- EM Nerd-The Case of the Partial Cohort - May 24, 2020
- EM Nerd: The Case of the Sour Remedy Continues - January 20, 2020
- EM Nerd-The Case of the Adjacent Contradictions - December 23, 2019