Kidney Injury Is Not Prevented By Hydration Alone

Jay Koyner; George Bakris


Eur Heart J. 2019;40(38):3179-3181. 

Acute kidney injury (AKI) is a common and important clinical syndrome that often occurs in the peri-operative and peri-interventional period (regardless of the technique employed) and has been specifically associated with increased morbidity and mortality[1] AKI in the peri-procedural setting remains a complex and challenging problem, with limited interventions having shown promise. In this issue of the European Heart Journal, Arbel et al. report the results of a single-centre, prospective, randomized, double-blind, sham-controlled clinical trial to determine the ability of automated saline infusion matched to furosemide-induced diuresis to prevent AKI in patients undergoing transcatheter aortic valve implantation (TAVI).[2] In this setting, AKI is often multifactorial, with potential sources including radiocontrast exposure, hypotension, and cardiogenic shock, extensive pre-existing chronic kidney disease and exposure to other agents that may impact kidney function (e.g. blood products or other nephrotoxins), and no single intervention can account for all of these issues (Take home figure). Thus, it should not be surprising the Arbel and colleagues were unable to demonstrate a decrease in AKI event rates in their trial of 136 patients, henceforth referred to as REDUCE-AKI.[2]

Take home figure.

Numerous etiologies that have been implicated in generating acute kidney injury during or following a TAVI procedure

It is important to note that before the publication of REDUCE-AKI, observational data and a small single-centre open-labelled randomized trial demonstrated a benefit of forced diuresis in reducing the peri-TAVI AKI rates.[3,4] These trials, including REDUCE-AKI, all utilized the RenalGuard system (RenalGuard Solutions Inc., Milford, MA, USA). This system, which is widely available across Europe, consists of a urinary catheter which has its collection bag connected to the RenalGuard consul that infuses i.v. solutions (in REDUCE-AKI, normal isotonic saline) back to the patient in order to match their furosemide-augmented urine output. While prior studies demonstrated a reduction in AKI events with the use of this forced diuresis with the matched hydration technique, the REDUCE-AKI study did not. It is noteworthy that AKI events were more likely to occur in those receiving the active treatment compared with those in the sham arm, causing the study to be stopped early due to this signal for harm and futility (25% vs. 19.1%; P = 0.41).

It is important to acknowledge, however, a few limitations around AKI in this trial. First, the forced diuresis with furosemide eliminates the potential for urine output-based AKI in some of the cohorts. Additionally, based on the AKI-Network criteria, the Valve Academic Research Consortium-2 (VARC-2) definition of AKI does not state that all patients receiving renal replacement therapy are considered to have stage 3 AKI. Specifically, there were five such patients in the REDUCE-AKI study, two in the active arm and three in the sham arm; P = 1.0. However, the authors report that the total number of severe AKI events (stage 2 or higher) was only three (one in the active group and two in the sham, P = 0.6). Thus, clarification is needed here.

Irrespective of this numerical discrepancy, the severe AKI event rates were low, and this was despite the high pre-procedure risk characteristics of the entire cohort. Patients in REDUCE-AKI mirrored those of previously published investigations demonstrating a benefit to the forced diuresis-matched hydration prevention strategy. They were, on average, 84 years of age, with nearly 40% having diabetes and a baseline estimated glomerular filtration rate (eGFR) just below 60 mL/min/m2. Thus, despite their Society of Thoracic Surgeon (STS) risk score being slightly lower than in prior published cohorts, the AKI event rates were slightly higher than prior publications (30 of 136; 22% of the entire cohort).

This increase in AKI events is even more surprising when one considers the impact of forced hydration on the evaluation and determination of kidney function using serum creatinine as the basis of the AKI diagnosis. In REDUCE-AKI, patients in the sham arm received i.v. normal saline at a rate of 0.5–1 mL/kg/h during and up to 6 h after the procedure. Those in the active group received a 250 mL bolus of saline followed by an i.v. bolus of furosemide (0.25–0.50 mg/kg). Once an active arm patient's urine output was above 300 mL/min, they underwent their procedure, with matched saline hydration continuing for 6 h after their last contrast dose of the catheterization procedure. This led to a mean difference of >3 L of i.v. saline administration in the active group compared with the sham group. Thus, given the large fluid loads in the active group, even in the setting of increased urine output, it is possible that this may have contributed to a haemodilution effect on serum creatinine, leading to misclassification and perhaps underestimation of the true AKI event rates. The impact of fluid balance on AKI misclassification is well established from the acute lung injury literature and may have impacted event rates in both arms of the REDUCE-AKI study.[5]

Another factor that could have contributed to the negative result was that all patients in both arms received normal saline. While there is some evidence that saline-based matched hydration with forced diuresis may improve patient outcomes,[4] there is equal mounting evidence that isotonic hyperchloraemic 0.9% saline is associated with increased risk of severe AKI and other major adverse kidney events in critically ill patients when compared with those receiving more physiological (balanced) i.v. fluids.[6] While a recent large international 2 × 2 factorial designed study demonstrated no difference in radiocontrast-associated AKI in patients undergoing angiography who received i.v. normal saline compared with those receiving 1.26% sodium bicarbonate,[7] it is important to highlight the fact that not all AKI post-TAVI is due to radiocontrast exposure.

While there may be a reason to explore the potential benefits of hydration and forced diuresis in the setting of TAVI utilizing a more physiological, balanced solution than normal saline, it is important to reflect on the aforementioned multifactorial nature of AKI in the peri-procedure setting. As such, it is unrealistic to expect a single intervention to cover all potential sources of AKI in a complex patient. To further exemplify this point, there have been several recent investigations that demonstrate enhanced AKI care and improved patient outcomes in high-risk patients using a multi-pronged AKI care bundle.[8,9] Focusing on more than just 'flushing out the kidney', these studies demonstrate that monitoring volume status and optimizing haemodynamics, avoiding unnecessary potentially nephrotoxic agents, ensuring euglycaemia, and close monitoring of serum creatinine and urine output coupled with early nephrology care leads to improved patient outcomes.

While some of this prior work is predicated on the use of novel biomarkers associated with AKI, the current tenants behind the primary prevention of AKI are not novel. To be clear, ensuring patients are euvolaemic and that the kidney receives the appropriate perfusion is part of the preventive strategy. Failure to achieve this helps explain some of the prior successes of forced diuresis-matched hydration strategies, but as previously noted there is more that can be done to mitigate peri-TAVI overall AKI risk, independent of radiocontrast exposure. Moreover, it remains possible that using this hydration–diuresis strategy, perhaps with balanced solutions, in conjunction with other functional and biochemical biomarkers of AKI as well as the AKI care bundle, may improve patient outcomes. However, this remains untested and, until these avenues are explored, there is no role for forced diuresis-matched hydration strategies to prevent AKI in the peri-TAVI period.