Skip to main content

Foundational Trials

PROTECTION

Image: Shutterstock / ChaNaWiT

Publication

  • Title: A Randomized Trial of Intravenous Amino Acids for Kidney Protection
  • Acronym: PROTECTION
  • Year: 2024
  • Journal published in: The New England Journal of Medicine
  • Citation: Landoni G, Monaco F, Ti LK, et al. A randomized trial of intravenous amino acids for kidney protection. N Engl J Med. 2024;391:687-698.

Context & Rationale

  • Background
    Cardiac surgery–associated acute kidney injury (AKI) is common and associated with increased morbidity, mortality, and subsequent chronic kidney disease; beyond supportive measures, proven preventive pharmacotherapy has been lacking.
  • Research Question/Hypothesis
    Does perioperative continuous intravenous infusion of a balanced amino acid mixture (intended to recruit renal functional reserve and increase renal perfusion) reduce postoperative AKI compared with placebo in adults undergoing elective cardiac surgery with cardiopulmonary bypass?
  • Why This Matters
    If a simple, low-complexity infusion can reduce AKI after cardiac surgery, the intervention is readily implementable at scale and may mitigate downstream kidney-related morbidity in a very large surgical population.

Design & Methods

  • Research Question: In adults undergoing elective cardiac surgery requiring cardiopulmonary bypass, does perioperative IV amino acid infusion reduce AKI (KDIGO creatinine criteria, stage ≥1 within 7 days) versus placebo?
  • Study Type: Investigator-initiated, multinational, multicentre (22 centres, 3 countries), randomised, double-blind, placebo-controlled, parallel-group trial; perioperative operating room + ICU setting.
  • Population:
    • Adults (≥18 years) scheduled for elective cardiac surgery requiring cardiopulmonary bypass; expected ICU stay ≥1 night; baseline creatinine available preoperatively.
    • Key exclusions included planned/current kidney-replacement therapy, and advanced chronic kidney disease (eGFR <30 mL/min/1.73 m2).
  • Intervention:
    • Blinded continuous infusion of balanced amino acid mixture (Isopuramin 10%) at 2 g/kg ideal body weight/day (maximum 100 g/day), started at operating-room admission.
    • Continued until 72 hours after initiation, ICU discharge, initiation of kidney-replacement therapy, or death (whichever occurred first).
    • If enteral/parenteral nutrition started within 72 hours, infusion rate adjusted to target total amino acid intake of 2 g/kg/day.
  • Comparison:
    • Blinded continuous infusion of placebo (Ringer’s solution) at the same rate and using the same stopping rules and adjustment approach.
  • Blinding: Double-blind: patients, clinicians, investigators, data collectors, and outcome assessors blinded; statisticians and authors blinded during analysis; pharmacists/trial nurses aware but not involved in data collection/analysis.
  • Statistics: Power calculation: planned n=3500 (1750/group) to detect a 20% relative risk reduction in AKI (assumed 25% in placebo) with 90% power at two-sided α=0.05, accounting for three interim analyses using O’Brien–Fleming boundaries; primary analyses by intention-to-treat (with prespecified missing-data imputations; plus per-protocol and as-treated analyses).
  • Follow-Up Period: Primary outcome assessed within 7 days after surgery; follow-up for mortality at ICU/hospital discharge and at 30, 90, and 180 days; quality of life at 180 days.

Key Results

This trial was not stopped early. Prespecified interim analyses were performed after enrolment of 25%, 50%, and 75% of the planned sample.

Outcome Amino acid group Placebo group Effect p value / 95% CI Notes
In-hospital AKI (KDIGO creatinine criteria), within 7 days 474/1759 (26.9%) 555/1752 (31.7%) RR 0.85 95% CI 0.77 to 0.94; P=0.002 Primary outcome; 2 missing in amino acid group.
Stage 3 AKI 29/1759 (1.6%) 52/1752 (3.0%) RR 0.56 95% CI 0.35 to 0.87; P not reported Lower severe AKI; stage 1 predominated overall (430 vs 492).
Kidney-replacement therapy during hospital stay 24/1759 (1.4%) 33/1752 (1.9%) RR 0.73 95% CI 0.43 to 1.22; P not reported Event rates low; no clear between-group difference.
Duration of kidney-replacement therapy (hours) Median 64 (IQR 17–100) Median 64 (IQR 21–81) Mean diff 9.79 hours 95% CI −62.13 to 81.70; P not reported Reported as absolute mean difference.
ICU length of stay (hours) Median 30 (IQR 21–66) Median 34 (IQR 21–68) Mean diff 1.39 hours 95% CI −4.78 to 7.57; P not reported Reported as absolute mean difference.
Hospital length of stay (nights) Median 7 (IQR 5–9) Median 7 (IQR 6–9) Mean diff 0.28 nights 95% CI −0.36 to 0.92; P not reported Reported as absolute mean difference.
30-day mortality 50/1759 (2.8%) 49/1752 (2.8%) RR 1.02 95% CI 0.69 to 1.50; P not reported No mortality signal despite AKI reduction.
Any adverse drug reactions 0 0 — Not applicable No ADRs reported in either group.
Surgical revision for bleeding 70/1759 (4.0%) 62/1752 (3.5%) RR 1.13 95% CI 0.81 to 1.58; P=0.48 Prespecified safety outcome.
Sepsis 75/1759 (4.3%) 76/1752 (4.3%) RR 0.98 95% CI 0.72 to 1.34; P=0.91 Prespecified safety outcome.
  • AKI occurred in 26.9% with amino acids vs 31.7% with placebo (RR 0.85; 95% CI 0.77 to 0.94; P=0.002).
  • Severe AKI (stage 3) was lower with amino acids: 1.6% vs 3.0% (RR 0.56; 95% CI 0.35 to 0.87).
  • Patient-centred secondary outcomes (kidney-replacement therapy, length of stay, mortality) were similar between groups.

Internal Validity

  • Randomisation and allocation: Central, web-based randomisation with computer-generated permuted blocks, stratified by site; concealment maintained by central service.
  • Blinding: Double-blind across clinicians/patients/outcome assessors; small numbers of unblinding by mistake (2 vs 2) reported.
  • Dropout/exclusions: One placebo-group patient withdrew consent before infusion initiation; otherwise near-complete follow-up for in-hospital outcomes.
  • Protocol adherence: Major deviations were uncommon: study infusion never started (16 [0.9%] vs 17 [1.0%]); premature and permanent discontinuation of infusion (48 [2.7%] vs 56 [3.2%]).
  • Crossover: Rare (3 [0.2%] vs 1 [0.1%]).
  • Baseline balance: Groups were closely matched (e.g., median age 66 vs 67 years; female 30.1% vs 30.1%; median preoperative creatinine 0.96 vs 0.94 mg/dL; hemofiltration during CPB 9.9% vs 9.2%).
  • Timing and dose: Infusion began at operating-room admission and continued until ICU discharge/72 hours/KRT/death; achieved similar infusion rates (median 40 mL/h in both groups).
  • Separation of the variable of interest: Median infused dose 1260 mL (126 g) vs 1272 mL (127 g); median duration 30 hours vs 31 hours; most stopped at ICU discharge (73.0% vs 71.1%).
  • Outcome assessment: Primary outcome based on KDIGO creatinine criteria; creatinine measurements prespecified up to 7 days (as feasible), with sensitivity analyses for missing values and urinary criteria reported.
  • Statistical rigour: Prespecified interim analyses; primary ITT with prespecified imputation approaches and supportive per-protocol/as-treated analyses showing consistent direction of effect.

Conclusion on Internal Validity: Strong. Allocation concealment and blinding were robust, event ascertainment was objective, deviations/crossover were infrequent, and the intervention exposure was well matched except for amino acid content.

External Validity

  • Population representativeness: Adult elective cardiac surgery with cardiopulmonary bypass in high-income settings; predominantly White (98%+), with 30% women; centres largely Italian with small numbers in Singapore and Croatia.
  • Key exclusions affecting generalisability: Advanced CKD (eGFR <30 mL/min/1.73 m2) and patients already receiving/planned for kidney-replacement therapy; results may not generalise to emergent cardiac surgery or off-pump pathways.
  • Applicability: Intervention is pragmatic (standard infusion product, bedside titration rules) and could be implemented where perioperative ICU infusion protocols exist; transferability to lower-resource settings is less certain given case-mix, baseline risk, and care pathways.

Conclusion on External Validity: Findings are highly applicable to elective on-pump cardiac surgery populations in high-income systems, but extrapolation to advanced CKD, emergency surgery, and settings with markedly different perioperative care infrastructure should be cautious.

Strengths & Limitations

  • Strengths: Large, multicentre, double-blind randomised design; objective primary outcome; prespecified interim analyses and sensitivity analyses; low crossover; consistent effect across supportive analyses.
  • Limitations: AKI defined primarily by creatinine (limited urine-output data capture post catheter removal); no tubular injury biomarkers; management bundle adherence (e.g., avoidance of hypotension/nephrotoxins) not reported; limited ethnic and geographic diversity; advanced CKD excluded.

Interpretation & Why It Matters

  • Clinical meaning
    Perioperative amino acid infusion reduced creatinine-defined AKI and stage 3 AKI after elective cardiac surgery, without detectable safety trade-offs, but did not change kidney-replacement therapy, length of stay, or mortality in this trial.
  • Mechanistic framing
    The pattern is compatible with recruitment of renal functional reserve (higher perfusion/filtration) rather than clearly demonstrated mitigation of tubular injury, given the absence of injury biomarker data.
  • Practice implication
    For centres seeking a pragmatic AKI-reduction strategy in elective on-pump cardiac surgery, amino acid infusion is a plausible option; whether it improves patient-centred outcomes or long-term kidney health remains uncertain.

Controversies & Subsequent Evidence

  • Functional AKI vs tubular protection: Absence of kidney injury biomarkers (e.g., NGAL, KIM-1) limits inference on whether the observed reduction in creatinine-defined AKI reflects true tubular protection or predominantly functional changes in filtration. 1
  • Baseline creatinine ascertainment and AKI definition: Concerns were raised that using preoperative creatinine values drawn from records (potentially up to 365 days before admission) could affect AKI classification; authors stated baseline creatinine was obtained in the week before surgery in all patients. 2
  • Placebo choice and chloride exposure: Criticism that the control fluid could differentially affect renal perfusion via chloride load was countered by authors noting low chloride concentration and low infusion rate, arguing physiological implausibility of chloride-driven effects. 2
  • Generalisation to advanced CKD / albuminuria: Exclusion of eGFR <30 mL/min/1.73 m2 and lack of proteinuria/albuminuria stratification were raised as limitations for broader kidney-risk phenotyping; authors argued concealed randomisation makes major baseline imbalances in such factors unlikely. 2
  • Subsequent evidence synthesis: A 2024 perioperative meta-analysis (including cardiac surgery trials) reported reduced AKI with IV amino acids overall and suggested reduced hospital length of stay, supporting a class effect on perioperative AKI risk, while still leaving uncertainty about long-term kidney outcomes and patient-centred endpoints. 3
  • Protocol transparency: The trial’s methodological manuscript and prespecified analytical approach were published before completion, supporting interpretability and reducing analytic flexibility concerns. 4

Summary

  • In 3511 adults undergoing elective on-pump cardiac surgery, perioperative amino acid infusion reduced AKI (26.9% vs 31.7%; RR 0.85; 95% CI 0.77 to 0.94; P=0.002).
  • Stage 3 AKI was also lower (1.6% vs 3.0%; RR 0.56; 95% CI 0.35 to 0.87).
  • Kidney-replacement therapy and patient-centred outcomes (length of stay, mortality) were similar between groups.
  • Protocol deviations and crossover were uncommon; blinding and allocation concealment were robust.
  • Absence of tubular injury biomarkers and exclusion of advanced CKD limit mechanistic inference and generalisation to highest-risk kidney phenotypes.

Further Reading

Other Trials

Systematic Review & Meta Analysis

Observational Studies

Guidelines

Notes

  • The primary outcome was defined using KDIGO creatinine criteria; a sensitivity analysis incorporating urinary criteria produced a similar relative risk estimate (reported by the trialists).

Overall Takeaway

PROTECTION is a large, double-blind perioperative RCT showing that a protocolised amino acid infusion modestly reduces creatinine-defined AKI and halves stage 3 AKI after elective on-pump cardiac surgery, without clear effects on kidney-replacement therapy or mortality. It advances the field by providing high-level evidence for a pragmatic, readily deployable AKI-prevention strategy, while leaving open whether benefits reflect functional changes in filtration versus true tubular protection and whether longer-term kidney outcomes improve.

Overall Summary

  • Perioperative IV amino acids reduced AKI (26.9% vs 31.7%) and stage 3 AKI (1.6% vs 3.0%) after elective on-pump cardiac surgery, with no clear differences in kidney-replacement therapy, length of stay, or mortality.

Bibliography