EDITORIAL

Amino Acid Infusion to Protect Kidney Function after Cardiac Surgery

Marlies Ostermann, Andrew D. Shaw

N Engl J Med 2024;391:759-760

DOI: 10.1056/NEJMe2408632

Cardiac surgery–associated acute kidney injury (AKI) is linked to adverse outcomes including an increased risk of death, postoperative complications, and higher rates of incident or progressive chronic kidney disease.^1,2 The cause of cardiac surgery–associated AKI is multifactorial. It is also well recognized that hemodynamic alterations can result in “functional” AKI, in which the glomerular filtration rate (GFR) decreases acutely without clinically significant tubular cell injury. In this latter context, the risk of adverse outcomes is lower.

In 1983, the administration of a protein load was shown to increase GFR. This finding introduced the concept of renal functional reserve,³ which is the capacity of the kidney to compensate or increase its function in states of either high metabolic demand or intrinsic kidney disease; thus, renal functional reserve represents increased renal plasma flow mediated by renal vasodilatation, predominantly within the preglomerular arteriolar vasculature.⁴ It has been suggested that recruitment of renal functional reserve may be possible through the administration of a protein load and that administering a protein load might be renoprotective.⁵ In a trial involving patients undergoing cardiac surgery in Japan, amino acid infusion increased postoperative estimated GFR (eGFR) and potentially reduced the risk of cardiac surgery–associated AKI.⁶ However, that trial included only 66 patients, and the incidence of AKI in the control group was surprisingly high at 56%.

In this issue of the Journal, Landoni et al. report the results from the Intravenous Amino Acid Therapy for Kidney Protection in Cardiac Surgery (PROTECTION) trial, which showed that an infusion of amino acids had a protective effect on kidney function after cardiac surgery.⁷ A total of 3511 patients with a baseline eGFR of at least 30 ml per minute per 1.73 m² who were scheduled for cardiac surgery were randomly assigned to receive an infusion of either amino acids (Isopuramin 10%, Baxter) or placebo (Ringer’s solution, Baxter). The infusion was started preoperatively and continued for 72 hours or until the patient’s death, initiation of kidney-replacement therapy, or the patient’s discharge from the intensive care unit. The intention-to-treat analysis showed that AKI of any stage (according to serum creatinine level) developed within the first week after surgery in significantly fewer patients in the amino acid group than in the placebo group. The incidence of stage 3 AKI was also lower in the amino acid group. The per-protocol analysis showed no significant difference in the occurrence of stage 2 or stage 3 AKI between the two groups, although this finding was driven by protocol violations that involved early or permanent interruption in the trial infusion, which would have been expected to favor placebo.

The PROTECTION trial has some clear strengths, such as a rigorous double-blind design and generalizability arising from recruitment of patients in 22 centers in three countries. In addition, the transparency in reporting is highly commendable. The strength of support provided by the prespecified subgroup analyses and the consistency of effect across study sites and sensitivity analyses are also reassuring for the top-line study results.

Positive results from AKI-prevention trials have been rare. So, are the results of the PROTECTION trial strong enough to translate into recommendations for including amino acid infusions in clinical practice? There are some issues to consider. First, 336 patients underwent hemofiltration during cardiac bypass as part of routine practice, which complicated the evaluation of postoperative creatinine levels. Second, the trial infusion was ceased prematurely in 104 patients, which would have introduced bias against the intervention and thus should not be viewed as a threat to the validity of the treatment effect. Third, as recognized by the authors, an important limitation is the lack of data on actual tubular injury, because levels of biomarkers that indicate renal tubular damage (damage biomarkers) such as neutrophil gelatinase–associated lipocalin or kidney injury molecule 1 are not reported. Evidence is accumulating that many of the consequences of cardiac surgery–associated AKI are more common in patients with overt tubular injury as defined by the presence of these and other damage biomarkers.⁸ Thus, whether the reduced incidence of AKI is a functional effect or whether there was true protection from tubular injury remains unclear. The reduction in the incidence of stage 3 AKI suggests that the treatment effect is real, but we note that there was no difference in the incidence of any outcomes that affect how patients function in terms of general health, including the use of kidney-replacement therapy; 30-day, 90-day, and 180-day mortality; median serum creatinine levels at these time points; and quality of life. Data on the percentage of patients with incident or progressive chronic kidney disease beyond day 90 were not reported, and a health economic analysis has not yet been conducted. Finally, adherence to general recommendations for the prevention of AKI postoperatively, including avoidance of hypotension and of the use of nephrotoxic drugs, was not reported.⁹

Positive results from large, well-conducted trials should drive practice change. The PROTECTION study is such a trial. However, whether the results are due to a functional effect or true protection against tubular injury, or a combination of both, remains unclear.