Postoperative 20% Albumin Infusion for Acute Kidney Injury in High-Risk Cardiac Surgery.

Shehabi et al.. JAMA Surgery 2025; doi:10.1001/jamasurg.2025.1683

 

Clinical Question

Amongst patients undergoing high risk cardiac surgery, does the administration of 20% Human Albumin Solution, compared to standard care, reduce the risk of post-operative acute kidney injury?

 

The primary outcome was the occurrence of stage 1-3 AKI (KDIGO criteria) up to 7 days after surgery.

Background

• Acute kidney injury occurs in close to 30% of individuals undergoing cardiac surgery, with an associated 3-28 fold increased risk of mortality and long-term renal dysfunction.
• Hypoalbuminaemia is observed in over 90% of cardiac surgery patients and correlates with AKI severity and mortality. It remains unclear whether albumin represents a modifiable therapeutic target or a marker of illness severity
• The original ALBICS trial (Lee et al, Anesthesiology 2016) demonstrated nephroprotective effects of 20% albumin when given before off-pump cardiac surgery, providing rationale for further investigation in on-pump surgery
• Recent guidelines highlighted the lack of evidence for albumin use in high-risk cardiac surgical patients and called for targeted trials (Callum et al, CHEST 2024)

Design

• Multicentre, open-label, parallel-group randomised controlled trial.
• 1:1 randomisation in permuted blocks with fixed block size, stratified by site and eGFR < 60 ml/min/1.73m².
• Open-label trial: patients and clinicians were unblinded.
• A modified intention-to-treat analysis was performed using generalised linear mixed models with binomial log link, adjusted for site and eGFR status.
• 620 participants required to give 80% power to detect a 10% reduction in the primary outcome, and an alpha = 0.05. The calculation assumed a 30% AKI rate in the control group and accounted for 5% drop-out.
• A pre-planned sensitivity analysis was performed to assess those patients in the control group who received > 500mL of 4% or 5% HAS.

Setting

• 7 cardiothoracic centres (6 Australian, 1 Italian) between July 2019 and August 2024.
• Recruitment was significantly impacted by COVID-19, with 90% of enrolment occurring between July 2022 and August 2024.

Population

• Inclusion criteria.

Adult patients having on-pump surgery, including either:

1. Combined cardiac surgery
2. Surgery on the thoracic aorta
3. Any cardiac surgery with eGFR < 60 ml/min/1.73m²

• Exclusion criteria
• ICU admission > 6 hours post-surgery
• eGFR < 15 ml/min/1.73m²
• Albumin < 20g/L
• Dialysis dependent or renal transplant recipient
• Off pump surgery
• ECMO or VAD recipients
• Objections to receiving albumin or blood products
• Participant numbers
• 954 patients screened, 621 randomised
• Modified Intention To Treat population:
  • 307 intervention, 304 control

• Patient characteristics.

Baseline characteristics were well balanced, representing contemporary high-risk cardiac surgical patients. Notable differences included higher diabetes rates in the control group (32% vs 27%) and higher AMI history (16% vs 14%).

Characteristic	20% Albumin (n=307)	Usual Care (n=304)
Age, mean (SD)	69.1 (11.0)	68.9 (10.6)
Male, %	70.7%	74.7%
Weight, mean (SD), kg	81.7 (19.0)	83.0 (19.1)
APACHE III score	50.7 (15.1)	51.8 (14.4)
EuroSCORE II, median (IQR)	3.16 (1.91-5.18)	3.30 (1.91-5.34)
Diabetes, %	27%	32%
Previous MI, %	14%	16%
eGFR <60 ml/min/1.73m², %	46%	46%
CPB time, median (SD), min	137 (49.6)	138 (56.4)
Cross-clamp time, mean (SD), min	105 (40.5)	106 (45.6)

• Comparison of perioperative data
• More Aortic surgeries in control group (33% vs 26%)
• CBP time very similar 137mins in intervention vs 138mins in control.
• Cross clamp time 105 mins in intervention vs 106 mins in control.
• End CPB haematocrit 30% in both groups.
• Similar intra-operative vasoactive and blood product administration.
• Cardiovascular supports at end of CPB well balanced: Noradrenaline / Adrenaline / Milrinone / IABP.

Intervention

• 20% Albumin Group: 300mL of 20% Human Albumin Solution at 20ml/hr over 15 hours, commenced within 6 hours of ICU admission.

Control

• Standard care at discretion of treating clinicians. No 20% albumin permitted for first 24 hours. Administration of 4% or 5% albumin permitted.

Management common to both groups

• Intraoperative management was similar between groups: comparable CPB times, cross-clamp times, and vasoactive support.
• Post-operatively, both groups received standard ICU care including crystalloid resuscitation, with iso-oncotic albumin (4-5%) permitted in the first 24 hours.
• Use of diuretics, initiation of renal replacement therapy and ventilation management and weaning, were all at the discretion of the treating clinicians.

Outcomes

• Primary Outcome: occurrence of KDIGO stage 1-3 AKI
  • 20% albumin increased the risk of AKI compared to usual care.

Analysis	20% Albumin	Usual Care	Adjusted RR (95% CI)	P value
Overall AKI	150/307 (48.9%)	132/304 (43.4%)	1.12 (1.04-1.21)	0.003
eGFR ≥60 ml/min/1.73m²	56/165 (33.9%)	52/165 (31.5%)	1.07 (0.83-1.38)	0.59
eGFR <60 ml/min/1.73m²	94/142 (66.2%)	80/139 (57.6%)	1.14 (1.07-1.22)	<0.001
Sensitivity analysis*			1.14 (1.01-1.30)	0.04

*Excluding control patients who received >500mL of 4% albumin

Statistically significant outcomes favouring usual care occurred in pre-specified subgroups including patients without diabetes, those with >2 hours CPB time, and those with mild or no LV dysfunction.

Sensitivity analysis

• Among those in usual care group who received > 500mLs of 4% HAS.
• Consistent with primary data:
  • Adjusted RR of 1.14 (1.01 – 1.3, p = 0.04).

Secondary Outcomes

• No statistically significant differences in
  • Major Adverse Kidney Events (MAKE) at 28 days.
  • Sustained AKI II or III.
  • RRT to D28
  • Hospital mortality at D28 (2% in intervention vs 0.7% in control)
  • Hospital and ICU LoS.
  • Duration of mechanical ventilation.
  • Vasopressor / inotrope free days at D14,

Tertiary outcomes

• Serum albumin at 24 hrs.
  • At the end of the first 24 hours, serum albumin levels were higher in the albumin group with a mean (SD) of 37.4 (8.4) vs 32.1 (6.5) g/L in the usual care groups, respectively (P < .001).
• Blood transfusion requirements at D2.
  • Higher rates of post-operative blood transfusion in intervention arm (38% vs 30%, p = 0.04)
• No difference in:
  • Fluid balance at 24 hrs.
  • Fluid balance at D2.
  • Fluid overload.
  • Re-intubation.
  • ICU re-admission.
  • Arrythmia requiring CV compromise.

Authors’ Conclusions

In cardiac surgery patients at increased risk of AKI, a continuous infusion of 20% albumin for 15 hours after surgery did not confer significant benefits and may increase the risk of post-operative AKI, particularly in patients with eGFR < 60 ml/min/1.73m².

Strengths

• Rigorous, pragmatic randomised clinical trial addressing genuine clinical equipoise
• Strong methodological design with pre-specified statistical analysis plan
• Good internal and external validity to Australian cardiothoracic ICUs
• Well-matched baseline characteristics and intraoperative management
• Robust findings consistent across sensitivity analyses and subgroup analyses
• Systematic review and meta-analysis by the same group supports conclusions

Weaknesses

• Open-label design permits potential treatment bias. Although a biochemically derived primary outcome is less prone to reporting and ascertainment biases.
• Primary outcome is biomarker-based rather than patient-centred
• AKI definition used creatinine criteria only, not urine output criteria
• Death as competing risk not statistically accounted for
• One centre contributed one-third of recruitment without statistical adjustment
• COVID-19 disruption prolonged recruitment timeline
• Late addition of Italian centre (10% of patients) with different treatment effect
• Usual care group received significantly greater volumes of iso-oncotic albumin, potentially confounding the results.

The Bottom Line

• 20% albumin infusion after high-risk cardiac surgery increases rather than prevents C-AKI, particularly in patients with pre-existing renal impairment (eGFR <60 ml/min/1.73m²)
• The significant increase in blood transfusion requirements (38% vs 30%) raises important questions about albumin’s impact on haemostasis and bleeding risk in cardiac surgery patients.
• These data do not support routine use of 20% albumin for AKI prevention and suggest clinicians should consider bleeding risk when considering albumin administration in cardiac surgical patients.
• Future research should investigate mechanisms underlying increased bleeding complications with albumin use in this high-risk population.

 

External Links

• article https://jamanetwork.com/journals/jamasurgery/fullarticle/2835041
• https://criticalcarereviews.com/meetings/ccr-down-under

Metadata

Summary author: Philip Emerson
Summary date: 7^th July 2025
Peer-review editor: Aniket Nadkarni