Kidney PROTECTION after Cardiac Surgery with Intravenous Amino Acid Infusion

AKI is a common complication after cardiac surgery.1–3 Although most cardiac surgery–associated AKI (CSA-AKI) is transient and mild, often mediated by postoperative hemodynamic lability, in some patients it may be severe and require support with acute KRT. CSA-AKI is also associated with significant short- and longer term morbidity and mortality, including development and progression of CKD. The pathogenesis of CSA-AKI has not been fully elucidated but is believed to be multifactorial, including hemodynamic alterations (including nonpulsatile blood flow during cardiopulmonary bypass [CPB] and low cardiac output states), ischemia-reperfusion injury, exposure to endogenous (e.g., free iron or heme pigments from hemolysis during CPB) and/or exogenous (e.g., medications) nephrotoxins, microemboli from vascular manipulation or the CPB circuit, neurohormonal activation, inflammation, complement activation, metabolic factors, and oxidative stress.1–3 Given the anticipated timing of exposure, CSA-AKI has been an appealing target for interventions to prevent or treat incipient kidney injury, including pharmacologic therapy with dopamine and dopaminergic agonists, diuretics, mannitol, natriuretic peptides, acetylcysteine, and statins; preemptive hemodialysis; adsorptive blood purification; remote ischemic preconditioning; blood filtration to capture atheromatous debris; and various strategies of goal-directed therapy. Unfortunately, none of these have been sufficiently efficacious for routine clinical use.1,2,4 One of the latest entries for prevention of CSA-AKI is the use of amino acid infusion to recruit renal functional reserve (RFR).5 RFR is the capacity of the kidney to increase GFR in response to increased demand, as occurs in response to increased protein intake.6 The observation that increased protein intake augments renal blood flow and glomerular filtration had been recognized for decades and was the basis of the Brenner hypothesis that elevated dietary protein results in glomerular hypertension and hyperfiltration, which in time is maladaptive and leads to progressive glomerulosclerosis.7 In the 1980s, Bosch et al. demonstrated that RFR could be recruited in humans with an acute protein load, documenting an increase in creatinine clearance in normal individuals from 110±13 to 171±8 ml/min 150 minutes after protein loading and observing a diminished response in individuals with decreased kidney function.6 The underlying mechanism by which protein ingestion or amino acid infusion alters renal hemodynamics and increases GFR remains incompletely understood, although contributory factors are believed to include direct effects on proximal tubular sodium reabsorption, inhibition of tubuloglomerular feedback, and local release of vasodilators including prostaglandins and nitric oxide.8,9 Decreased RFR has been associated, in a small number of patients, with increased risk of development of or progression of CKD.9 Similarly, RFR has been proposed as a biomarker of risk of CSA-AKI. In a series of 110 patients undergoing cardiac surgery, RFR was 15.5±8.7 ml/min per 1.73 m2 in 15 patients who developed postoperative AKI as compared with 27.0±8.6 ml/min per 1.73 m2 among those who did not develop AKI (P < 0.001) despite similar baseline GFR.10 The Intravenous Amino Acid Therapy for Kidney Protection in Cardiac Surgery (PROTECTION) trial tested the hypothesis that recruitment of RFR using postoperative intravenous amino acid infusion would decrease CSA-AKI.5 This multinational trial randomized 3512 individuals with an eGFR of at least 30 ml/min per 1.73 m2 scheduled to undergo elective on-pump cardiac surgery and who were expected to require intensive care unit (ICU) care for at least one night to receive an infusion of either 2 g/kg per day of amino acids or a matched infusion of Ringer's solution beginning at operating room admission and continuing for up to 72 hours (or until ICU discharge, initiation of KRT, or death). The primary outcome was occurrence of AKI, as defined by an increase in serum creatinine of ≥0.3 mg/dl over <48 hours or an increase of ≥50% over 7 days, with secondary outcomes including AKI stage, use of KRT, and mortality through 180 days. A total of 1759 participants were randomized to receive the amino acid infusion as compared with 1752 who received placebo infusion and were well matched with regard to baseline demographics, comorbidities, cardiovascular status, and intraoperative management. Median preoperative serum creatinine was 0.96 mg/dl (interquartile range, 0.81–1.11 mg/dl) in the amino acid group compared with 0.94 mg/dl (interquartile range, 0.81–1.10 mg/dl) in the control group. 9.9% in the amino acid group received intraoperative hemofiltration during CPB as compared with 9.2% in the control arm. Protocol adherence was comparable in the two treatment arms. The key finding was a 15% lower rate of AKI among patients assigned to the amino acid infusion than among individuals assigned to placebo (26.9% versus 31.7%; relative risk, 0.85; 95% confidence interval, 0.77 to 0.94; P = 0.002). Most AKI episodes were mild (stage 1); however, 1.6% of individuals who received amino acids as compared with 3.0% of those in the control arm (relative risk, 0.56; 95% confidence interval, 0.35 to 0.87) developed stage 3 AKI. Rates of KRT were 1.4% and 1.9%, respectively. There was no significant difference in mortality before ICU discharge or at day 30, 90, or 180 or in kidney function, with a median serum creatinine at day 180 of 0.94 mg/dl in both treatment arms. How should we interpret these results? Should amino acid infusions to recruit RFR become standard therapy to prevent CSA-AKI? Although the concept of using interventions that increase kidney perfusion and GFR is attractive, other interventions to augment intrarenal hemodynamics, including use of kidney vasodilators and natriuretic peptides, have not been beneficial.1–3 The predominant benefit seen in the PROTECTION trial was in ameliorating small, transient changes in kidney function that are unlikely to reflect significant intrinsic kidney damage. Unfortunately, no biomarkers of tubular injury were collected to allow phenotyping of AKI. While the authors argue that biomarkers of injury have not been histologically validated in CSA-AKI, their elevation would have suggested that the creatinine increase was more than a transient hemodynamic effect and was associated with structural damage. Notably, there was also no benefit in either mortality over the first 6 months postoperatively or kidney function at 30, 90, or 180 days postoperatively. The largest concern in interpreting the potential benefit of RFR recruitment is related to its direct effect on the primary end point, which was assessed on the basis of change in serum creatinine. The increased glomerular filtration from recruitment of RFR directly results in a reduction in serum creatinine. In both treatment arms, serum creatinine values declined by approximately 0.1 mg/dl in the immediate postoperative period, most likely reflecting intraoperative volume administration, but then increased over the next 2 days, before returning to baseline. As would be expected, however, from an intervention that increases GFR, the rise in serum creatinine among the individuals receiving amino acid infusion was smaller than among controls, by approximately 0.05 mg/dl at day 2. This systematic lowering of serum creatinine by the intervention may seem trivial but is sufficient to diminish the apparent incidence of AKI as defined by small changes in serum creatinine (Figure 1). Thus, the apparent reduction in the incidence of AKI in the PROTECTION trial may be more an ephemera than a meaningful clinical benefit. Similar caution must be exercised in interpreting the putative benefits of multiple other interventions for prevention of AKI that may alter serum creatinine independent of actual benefit in preventing intrinsic kidney injury.Figure 1: Effect of intervention that systematically reduces serum creatinine concentration on apparent incidence of AKI. (A) Distribution of increase in serum creatinine after an event (e.g., postcardiac surgery) associated with development of AKI, where the x axis represents the change in serum creatinine from baseline and the height of the curve (y axis) the percentage or number of individuals sustaining the change. The black vertical broken line represents no change in serum creatinine. An arbitrary level of change in serum creatinine (e.g., an increase of ≥0.3 mg/dl over 48 hours or an increase of ≥50% over 7 days) is defined as AKI (cross-hatched area). (B) If an intervention systematically lowers the serum creatinine concentration (e.g., recruitment of RFR), the distribution curve for change in serum creatinine will be shifted to the left (blue curve). Using the same definition of AKI, the incidence of AKI is reduced (solid red area) without any actual protection of the kidney from intrinsic injury. (C) For comparison, the pattern of change for a hypothetical intervention that provides nephroprotection without systematically lowering the serum creatinine concentration (green curve) is shown. The incidence of AKI (solid orange area) is similarly reduced as in (B) but without an overall leftward shift of the distribution curve of the change in serum creatinine. RFR, renal functional reserve.