Trial by Combat: Do Amino Acids provide PROTECTION to the Kidneys?

More information
Ensayo Clínico PROTECTION - El Resumen Visual
Ensayo Clínico PROTECTION - El Resumen Visual
PROTECTION trial - visual abstract
PROTECTION trial - visual abstract
Trial by Combat: Do Amino Acids provide PROTECTION to the Kidneys?
Trial by Combat: Do Amino Acids provide PROTECTION to the Kidneys?

This week, we will discuss an intriguing hypothesis for many nephrologists: can amino acids prevent acute kidney injury?

Introduction

Acute Kidney Injury (AKI) is a common problem amongst hospitalized patients, particularly if they are undergoing surgery, cardiac surgery for example. The incidence of cardiac surgery-associated AKI is around 23-50%, and 2%–5% of these patients require kidney replacement therapy (KRT) during an AKI episode (D’Onofrio A, et al, Congest Heart Fail. 2010, Lagny MG, et al, BMC Nephrol. 2015). There is also the potential for development of CKD, rehospitalization and higher mortality following an AKI event. Many different strategies have been tried to protect the kidneys from AKI, but nothing has shown any effect so far. There are many hemodynamic factors that contribute to the pathophysiology of cardiac surgery-associated AKI. Hypoperfusion of the kidneys plays an important role in AKI and the insult is often worsened by the presence of underlying chronic kidney disease.

 Pathophysiology of cardiac surgery–associated acute kidney injury, from Sreekanth R, et al, Anesthesiology 2023.

A large network metanalysis of randomized controlled trials evaluating strategies for post–cardiac surgery AKI prevention only showed a small number of potentially beneficial strategies. (Chen J J, et al, Front Cardiovasc Med, 2022) Out of the 161 trials and 53 strategies evaluated, only 9 were suggested to be beneficial for cardiac surgery associated AKI, all nine potentially beneficial strategies were supported by only moderate- to low-confidence evidence. Note that amino acid was one of the studied strategies and it did not have a statistically significant result.

Specifically with respect to amino acids, previously, only small trials have been reported. For example, in 2019 Pu H et. al. The Journal of Thoracic and Cardiovascular Surgery published a single-center, randomized trial, concluding that IV amino acid therapy for kidney protection did not alter duration of renal dysfunction. In addition, clinical characteristics of the control group tend to be reported with longer cardiopulmonary bypass and operation times. Also, as expected, we can see an elevation in blood urea nitrogen levels. (Kazawa M et. al. JA Clin Rep, 2024) Despite these small negative trials, it was thought that a larger definitive trial might produce different results, which lead to the Intravenous Amino Acid Therapy for Kidney Protection in Cardiac Surgery (PROTECTION) trial. 50 years after the journal published a small RCT (Abel et al, NEJM 1973) showing the implausible effects of IV amino acids post AKI in recovery, the topic is now being revisited in a really large trial.

The Study

Methods

The PROTECTION trial was a multinational, double-blind, randomized, placebo-controlled trial conducted at 22 centers in three countries (Italy, Singapore, Croatia). The purpose of the trial was to study whether intravenous amino acid therapy would lead to a lower occurrence of postoperative AKI.

Study population

Inclusion criteria

  • Adults >18 years or older

  • Elective cardiac surgery requiring cardiopulmonary bypass

  • Expected to stay in ICU for at least one night after surgery

Exclusion criteria 

  • Currently enrolled into another randomized clinical trial

  • Currently receiving or scheduled for intermittent or continuous kidney-replacement therapy

  • CKD equal or more than stage 4 (authors state as eGFR <30 ml/min/1.73 m2 estimated by Cockroft-Gault equation - which estimates creatinine clearance in ml/min and not eGFR, so read it as such throughout the manuscript) or a kidney transplant

  • Not expected to survive ICU or hospital discharge

  • Severe liver disease (Child-Pugh score >7 points)

  • Hypersensitivity to one or more of the included amino-acids

  • Congenital alteration of amino acid metabolism

  • Pregnant or currently breastfeeding patients

  • Patients with any of the contraindications reported in the summary product characteristics

All eligible patients were required to have a baseline creatinine measured before surgery, obtained either during this hospitalization for surgery or within 365 days before this hospitalization.

Inclusion and exclusion criteria. Adapted from: Landoni G, et al. NEJM, 2024.

Interventions and follow-up

Patients were randomized to receive amino acid infusion or placebo based on a computer-generated, permuted-block sequence with stratification according to site. Trial oversight included an independent and data safety monitoring committee performing blinded interim analyses at 25%, 50%, and 75% enrollment stages. Patients, physicians, investigators, data collectors, and outcome assessors including statisticians and authors were blinded to the group assignments. Only the pharmacists and trial nurses were not blinded, but they were excluded from the data tasks. In the amino acid group, patients received a continuous infusion of a balanced mixture of amino acids through a central venous line, at a dose of 2 g per kilogram of ideal body weight per day (up to a maximum of 100 g per day) from the time of operating-room admission until 72 hours after initiation of the infusion, discharge from the ICU, initiation of kidney-replacement therapy, or death (whichever occurred first). Adjustments were made for concurrent nutrition to maintain the target amino acid intake. The amino acid solution had an osmolarity of 949 mOsm/L and a pH of 6-7. 

Amino Acid composition of Isopuramin 10%, supplement pg 25. Landoni G, et al. NEJM, 2024

In the placebo group, patients received a continuous infusion of Ringer’s solution with a rate calculated based upon what they would have needed to receive 2 g/kg of ideal body weight per day of the amino acid solution.

A blinded investigator performed phone follow-ups at 30, 90, 180 days. If the patients couldn’t be contacted, the data were obtained from their surgeon or general practitioner. 

Primary Outcome

The primary outcome was the occurrence of AKI within the first week after surgery. AKI was defined according to the KDIGO creatinine criteria, but not the urine output criteria, as specified in the initial published protocol (Landoni G, et al, Contemp Clin Trials, 2022, NCT03709264). Despite exclusion of urine output for AKI staging, it was diligently noted for RRT indication later in the results.

Secondary outcomes

  • Severity of AKI according to the KDIGO criteria

  • Use of kidney-replacement therapy during the hospital stay and the duration of kidney-replacement  therapy 

  • Duration of stay in the ICU and hospital

  • Duration of mechanical ventilation

  • Death from any cause documented at the time of ICU discharge, hospital discharge, or 30, 90, or 180 days after randomization 

  • Prespecified subgroup analyses according to age (≥75 vs <75 years of age), New York Heart Association class (I or II vs III or IV), and baseline eGFR (≤60 vs >60 ml per minute per 1.73 m2)

  • Prespecified intraoperative and postoperative adverse events 

Statistical analysis and sample size calculation

The authors hypothesized that AKI would develop in 25% of the patients in the placebo group based on previous studies. They estimated that a total of 3500 patients with 1750 patients in each arm would give the trial 90% power to detect a 20% lower risk of AKI in the amino acid infusion group than in the placebo group, at a two sided alpha level of 0.05 (accounting for the rounding and the three interim analyses). Interim analyses were conducted as planned, and primary analyses followed the intention-to-treat principle. Final data were not adjusted for multiplicity analysis.

Funding source

The PROTECTION trial was funded by the Italian Ministry of Health. One of the authors received a research grant from Baxter, with the same author being part of the steering committee. Baxter Inc. provided the amino acid treatment (Isopuramin 10%) and matching placebo (Ringer’s solution).

Results

Around 4415 patients were screened for eligibility during the study period from October 2019 through January 2024 and an impressive 3512 (80%) were enrolled, of which one patient withdrew consent before initiation of the trial regimen. Interestingly, 67 participants were excluded after consenting but before randomization, since they had hitherto unnoticed exclusion criteria. Finally, 1759 patients were randomly assigned to the amino acid group and 1752 to the placebo group. Additionally there were 183 protocol violations with 198 major protocol deviations (details in Table S3). 

Figure S1: Consort flow diagram. Landoni G, et al. NEJM, 2024

The median age of patients was ~ 66 years and only 30% of the patients were females. Patients in both groups were predominantly (98%) White, given the trial conducted in Italy and Croatia with only one site in Singapore presumably with the 2% non-White ethnicity. Requirements for intraoperative inotropes, vasoactive drugs and loop diuretics were similar in both groups. About 75% of  the study population in both arms had HF class II or III NYHA, with a median LVEF of approximately 60%. Less than 20% of the patients had diabetes, and only about 40% were on statins. About 35% underwent a CABG, with the rest being valvular surgery (and the percentages being >100% suggests some were CABG + valve surgery). 

Table 1: Baseline demographic characteristics. Landoni G, et al. NEJM, 2024 

Trial Interventions

The median infusion dose of amino acid was 126 grams, and the median infusion duration was 30 hours. The reasons for cessation of infusion included discharge from the intensive care unit (72.1%), initiation of kidney-replacement therapy (0.6%), discontinued in error (1.6%), death (0.4%), and withdrawal of consent (<0.1%). Only 22.6% of patients completed the maximum planned 72-hour infusion!

Table 2: Details of trial interventions.  Landoni G, et al. NEJM, 2024

There was no significant difference in blood pressure, central venous pressure (table S10), hematocrit (table S9) and urine output (table S11) between arms. Mentionable is that at day 2, approximately 70% of data for both of arms are lacking, concordant percentage with patients discharged from ICU. 

Table S10: Hemodynamic data in ICU. Landoni G, et al. NEJM, 2024

Primary Outcomes

The primary outcome was the occurrence of  AKI stages 1-3, which developed in 474 patients in the amino acid group (26.9%) and in 555 patients in the placebo group (31.7%), with a RR of 0.85; 95%CI 0.77 to 0.94. An overwhelming majority of AKI was stage 1 - 430 of 474 patients (90%) in the amino acid group and 492 of 555 (89%) in the placebo group. Stage 3 AKI, though rare (< 3% in both groups), was seen in 52 patients (3%) in the placebo group, but only 29 patients (1.6%) in the amino acid group (RR 0.56; 95% CI 0.35-0.87).

Table 3: Primary and Secondary Outcomes. Landoni G, et al. NEJM, 2024

Figure S2 shows a slightly lower (less than 0.05 mg/dl) difference in serum creatinine between the two arms, which was not statistically significant. 

Figure S2: Serum creatinine from baseline, in the first 72h. Landoni G, et al. NEJM, 2024

Secondary Outcomes

Kidney-replacement therapy was required in 24 (1.4%) patients in the amino acid group and 33 (1.9%) patients in the placebo group. The median duration of kidney-replacement therapy, mechanical ventilation, and duration of stay in hospital was similar between both groups. The 30-day and 180-day mortality rates were also similar between both groups. 

Subgroup analysis

Overall, the subgroup analysis just confirms the small numbers in the main analyses, but was overall consistent.

Figure S4: Planned subgroup analysis for primary outcome. Landoni G, et al. NEJM, 2024

Kidney-replacement therapy

There was no significant difference in kidney replacement therapy between the two groups- 1.4% vs 1.9%, only a difference of 9 patients (RR 0.73; 95%CI 0.43-1.22). The criteria for kidney replacement therapy initiation are given in table S12, and the most common indication was “severe oliguria” defined as urine output <100 ml over 6 hours.

Table S12: Criteria for Kidney-Replacement Therapy Initiation. Landoni G, et al. NEJM, 2024

A Kaplan–Meier time-to-event plot for kidney-replacement therapy is shown in S6. As expected, most AKI requiring KRT occurred within 5 days of surgery, with very few late occurrences. Interestingly, despite higher urea levels in the intervention arm (table S9) this was appropriately not used in RRT initiation indications. 

Figure S6: Kaplan-Meier Curve for Kidney-replacement Therapy. Landoni G, et al. NEJM, 2024

Safety events

There was no significant difference in prespecified adverse events including cardiogenic shock, events requiring ICU readmission, arrhythmia, need for reintubation, need for NIV, adverse neurologic outcomes, myocardial infarction, wound infection, sepsis, septic shock, and death between the two groups. The 30-day mortality was 2.8% (50 patients)in the amino acid group and 2.8% (49 patients) in the placebo group.

Table 4: Safety events in both study groups. Landoni G, et al. NEJM, 2024

Discussion

The common adage “Prevention is better than cure” is as relevant today as it was in the early 1500s, when first coined by the Dutch philosopher Erasmus. This sentiment was later paraphrased by Benjamin Franklin as “An ounce of prevention is worth a pound of cure”, although in this instance we are talking about mere grams of amino acids!  Amino acid infusions have been tested in preclinical and human studies since the 1990s. In the present trial, the authors report an impressive 5% absolute and 15% relative risk reduction in stage 1 AKI, as well as a smaller but significant 1.4% and 44% absolute and relative risk reduction in stage 3 AKI. However, the difference in need for RRT, mortality and other clinically relevant outcomes were not significant. Before we jump into implementing amino acid infusions, let’s try to understand how amino acids might work, and whether these results are robust.   

The proposed mechanism by which amino acid infusion protects the kidney from AKI is by improved renal medullary perfusion and improved glomerular blood flow. (Jeppsson A, et al, Journal of cardiothoracic and vascular anesthesia. 2000)  For every nephrologist it might sound counterintuitive, as for long time the mantra was that high dietary protein intake can cause intraglomerular hypertension, which may result in kidney hyperfiltration, glomerular injury, and proteinuria. (Ko GJ et al, JASN, 2020) This same physiological effect is postulated to be potentially beneficial in AKI. It was hypothesized that amino acid infusion increased eGFR by recruiting renal functional reserve (RFR), thus providing protection against AKI. RFR describes the capacity of the intact nephron mass to increase GFR from baseline in response to stimuli. Prior studies  have shown that RFR can predict the risk of AKI after cardiac surgery, hence recruitment of RFR, with help of interventions like high protein diet or amino acid infusion, might confer protection from AKI. (Husain-Syed F, et al, Ann Thorac Surg. 2018) 

Potential mechanisms by which amino acid infusion protects kidney function
Bio: Jufar Ah et al, Am J Physiol Regul Integr Comp Physiol, 2020, G. Landoni, et al. NEJM, 2024.

Limitations

Though nearly a third of patients in both groups (as expected for statistical analysis) developed AKI, almost 90% of those events were AKI stage 1, or minor with no significant differences in the clinically relevant outcomes of RRT, length of stay, or mortality. Finally, there was no difference between groups in the 30-day, 90-day and 180-day creatinine, suggesting no benefit of amino acid infusion on long term kidney function. A change in creatinine number is not exciting in the setting of a lack of effect on clinical outcomes. This is all the more relevant given the (likely) high cost of IV amino acids for this small and uncertain benefit. 

The change in creatinine should also be put into some more context. Less than a quarter of the participants received the complete 72 hour course of intervention. There was no change in any of the usual hemodynamic mechanistic pathways for AKI development. The lack of any data on tubular injury biomarkers adds to this opacity about how the amino acids actually had any effect.

The trial authors did not include a strict protocol for management or prevention of AKI in this trial but the use of KDIGO guidelines was recommended. The trial population was from high-income countries and differed substantially from patients in low and middle-income countries, affecting generalizability. Additionally, some patients had significant protocol deviations (Table S3), making the potential optimal dose and duration of amino acid infusion still unknown. Lots of chinks in the armor make this intervention to prevent AKI hard to justify, and unlikely to deflect even glancing blow to the kidneys.

Table S3: Protocol deviations. Landoni G, et al. NEJM, 2024. 

Conclusion

This trial reports that amino acid infusion in the perioperative period reduced acute kidney injury in patients undergoing cardiac surgery with cardiopulmonary bypass. Unfortunately, the main outcome was primarily reduction in early stages of AKI without decreases in hard clinical outcomes. Given the other significant limitations, amino acid infusions for AKI are unlikely to find a mainstream audience based upon the current study.

Summary prepared by 

Anjana Gopal,
Assistant Professor, Department of Nephrology, 
Amrita Institute of Medical Sciences, Kochi, Kerala, India

and

Samantha Verdugo, 
Internal Medicine Resident,
Guadalajara, Jalisco, Mexico

NSMC interns Class of 2024

Reviewed by Brian Rifkin, Cristina Popa, Sabarinath, and Swapnil Hiremath

Header Image created by AI, based on prompts by Evan Zeitler