Effect of 4% Albumin Solution vs Ringer Acetate on Major Adverse Events in Patients Undergoing Cardiac Surgery With Cardiopulmonary Bypass: A Randomized Clinical Trial

Eero Pesonen; Hanna Vlasov; Raili Suojaranta; Seppo Hiippala; Alexey Schramko; Erika Wilkman; Tiina Eränen; Kaapo Arvonen; Maxim Mazanikov; Ulla-Stina Salminen; Mihkel Meinberg; Tommi Vähäsilta; Liisa Petäjä; Peter Raivio; Tatu Juvonen; Ville Pettilä

doi:10.1001/jama.2022.10461

. 2022 Jul 19;328(3):251–258. doi: 10.1001/jama.2022.10461

Effect of 4% Albumin Solution vs Ringer Acetate on Major Adverse Events in Patients Undergoing Cardiac Surgery With Cardiopulmonary Bypass

A Randomized Clinical Trial

Eero Pesonen ^1,^✉, Hanna Vlasov ¹, Raili Suojaranta ¹, Seppo Hiippala ¹, Alexey Schramko ¹, Erika Wilkman ¹, Tiina Eränen ², Kaapo Arvonen ¹, Maxim Mazanikov ¹, Ulla-Stina Salminen ³, Mihkel Meinberg ¹, Tommi Vähäsilta ³, Liisa Petäjä ¹, Peter Raivio ³, Tatu Juvonen ³, Ville Pettilä ¹

¹Department of Anesthesiology, Intensive Care and Pain Medicine, University of Helsinki and Helsinki University Hospital, Helsinki, Finland

²HUS Pharmacy, University of Helsinki and Helsinki University Hospital, Helsinki, Finland

³Department of Cardiac Surgery, Heart and Lung Center, University of Helsinki and Helsinki University Hospital, Helsinki, Finland

^✉

Corresponding Author: Eero Pesonen, MD, PhD, Helsinki University Hospital, Kirurginen sairaala, PL 263, 00029 HUS, Helsinki, Finland (eero.pesonen@hus.fi).

Accepted for Publication: June 2, 2022.

Author Contributions: Dr Pesonen had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Pesonen, Vlasov, Suojaranta, Hiippala, Schramko, Wilkman, Arvonen, Mazanikov, Salminen, Vähäsilta, Juvonen, Pettilä.

Acquisition, analysis, or interpretation of data: Pesonen, Vlasov, Suojaranta, Hiippala, Wilkman, Arvonen, Mazanikov, Salminen, Meinberg, Vähäsilta, Petäjä, Raivio, Juvonen.

Drafting of the manuscript: Pesonen.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Pesonen.

Obtained funding: Pesonen.

Administrative, technical, or material support: Eränen.

Supervision: Pesonen, Suojaranta, Hiippala, Juvonen, Pettilä.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported by Prothya Biosolutions (formerly Sanquin Plasma Products), Amsterdam, the Netherlands, which as the manufacturer of the albumin solution products provided the solution and financial support, and by the research fund of Helsinki University Hospital.

Role of the Funder/Sponsor: Prothya Biosolutions participated in the design of pharmacovigilance surveillance; otherwise, it did not participate in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation or approval of the manuscript; and decision to submit the manuscript for publication. The sponsor participated in the management of the pharmocovigilance data and review of the manuscript. The sponsor did not have the right to veto publication or to control the decision regarding to which journal the manuscript was submitted.

Additional Contributions: We thank Markku Peltonen, PhD, National Institute for Health and Welfare, Helsinki, Finland, for designing the statistical analysis plan, review of the interim analysis, and conduct of the final statistical analysis and Aki Niemi, LicSocSC, Helsinki, Finland, for conduct of the interim statistical analysis, both of whom received compensation; Jouko Jalonen, MD, PhD, Department of Anesthesiology and Intensive Care, University of Turku, Turku, Finland, and Anders Perner, MD, PhD, Department of Intensive Care, Copenhagen University Hospital–Rigshospitalet, Copenhagen, Denmark, who also reviewed the interim analysis, for which they received no compensation; and study nurses Katri Meriläinen, RN, and Hanna Männynväli, RN, Department of Anesthesiology, Intensive Care, and Pain Medicine, University of Helsinki and Helsinki University Hospital, Helsinki, Finland, who participated in the study conduct and data acquisition, both of whom received compensation.

Data Sharing Statement: See Supplement 4.

^✉

Corresponding author.

PMCID: PMC9297113 PMID: 35852528

Key Points

Question

Does use of 4% albumin solution compared with Ringer acetate as cardiopulmonary bypass prime and perioperative intravenous volume replacement reduce the risk of major perioperative and postoperative complications in cardiac surgery?

Findings

In this randomized clinical trial that included 1386 patients undergoing cardiac surgery, the combined incidence of major perioperative and postoperative complications in the groups receiving 4% albumin vs Ringer acetate was 37.1% vs 33.8%, respectively. This difference was not statistically significant.

Meaning

These findings do not support the use of 4% albumin solution in cardiac surgery with cardiopulmonary bypass.

Abstract

Importance

In cardiac surgery, albumin solution may maintain hemodynamics better than crystalloids and reduce the decrease in platelet count and excessive fluid balance, but randomized trials are needed to compare the effectiveness of these approaches in reducing surgical complications.

Objective

To assess whether 4% albumin solution compared with Ringer acetate as cardiopulmonary bypass prime and perioperative intravenous volume replacement solution reduces the incidence of major perioperative and postoperative complications in patients undergoing cardiac surgery.

Design, Setting, and Participants

A randomized, double-blind, single-center clinical trial in a tertiary university hospital during 2017-2020 with 90-day follow-up postoperatively involving patients undergoing on-pump coronary artery bypass grafting; aortic, mitral, or tricuspid valve surgery; ascending aorta surgery without hypothermic circulatory arrest; and/or the maze procedure were randomly assigned to 2 study groups (last follow-up was April 13, 2020).

Interventions

The patients received in a 1:1 ratio either 4% albumin solution (n = 693) or Ringer acetate solution (n = 693) as cardiopulmonary bypass priming and intravenous volume replacement intraoperatively and up to 24 hours postoperatively.

Main Outcomes and Measures

The primary outcome was the number of patients with at least 1 major adverse event: death, myocardial injury, acute heart failure, resternotomy, stroke, arrhythmia, bleeding, infection, or acute kidney injury.

Results

Among 1407 patients randomized, 1386 (99%; mean age, 65.4 [SD, 9.9] years; 1091 men [79%]; 295 women [21%]) completed the trial. Patients received a median of 2150 mL (IQR, 1598-2700 mL) of study fluid in the albumin group and 3298 mL (IQR, 2669-3500 mL) in the Ringer group. The number of patients with at least 1 major adverse event was 257 of 693 patients (37.1%) in the albumin group and 234 of 693 patients (33.8%) in the Ringer group (relative risk albumin/Ringer, 1.10; 95% CI, 0.95-1.27; P = .20), an absolute difference of 3.3 percentage points (95% CI, −1.7 to 8.4). The most common serious adverse events were pulmonary embolus (11 [1.6%] in the albumin group vs 8 [1.2%] in the Ringer group), postpericardiotomy syndrome (9 [1.3%] in both groups), and pleural effusion with intensive care unit or hospital readmission (7 [1.0%] in the albumin group vs 9 [1.3%] in the Ringer group).

Conclusions and Relevance

Among patients undergoing cardiac surgery with cardiopulmonary bypass, treatment with 4% albumin solution for priming and perioperative intravenous volume replacement solution compared with Ringer acetate did not significantly reduce the risk of major adverse events over the following 90 days. These findings do not support the use of 4% albumin solution in this setting.

Trial Registration

ClinicalTrials.gov Identifier: NCT02560519

This randomized clinical trial assesses the safety and effectiveness of 4% albumin vs Ringer acetate as the priming solution of cardiopulmonary bypass and intraoperative and postoperative volume replacement in patients undergoing on-pump cardiac surgery.

Introduction

Optimal fluid therapy in cardiac surgery is controversial. Colloids may maintain hemodynamics better than crystalloids and may be more suitable for restrictive fluid therapy.^1,2 In addition to its colloid osmotic effect, albumin transports hydrophobic molecules, preserves the endothelial glycocalyx, has antioxidative and anti-inflammatory functions as well as possesses enzymatic and signaling activites.^3,4 In cardiac surgery, albumin solution in cardiopulmonary bypass priming (ie, fluid filling of the bypass tubes, oxygenator, and reservoir) was shown to maintain cardiac function better than crystalloids.² Furthermore, albumin in priming reduced the decrease in platelet count and excessive fluid balance.^2,5

The effects of albumin on kidney function in cardiac surgery are conflicting. In one randomized clinical trial (RCT) in on-pump cardiac surgery, serum creatinine levels were higher postoperatively among patients receiving albumin than among those receiving Ringer lactate.⁶ In another RCT of off-pump cardiac surgery, preoperative administration of 20% albumin increased urine output and reduced incidence of acute kidney injury compared with saline.⁷ Evidence about the hemostatic effects of albumin in cardiac surgery is scarce. In an RCT comparing 4% albumin, hydroxyethyl starch, and Ringer lactate in 240 patients undergoing on-pump cardiac surgery, postoperative chest tube drainage did not differ between the groups, but the patients in the albumin and hydroxyethyl starch groups received red blood cells and fibrinogen concentrate more frequently.⁶

The objective of the Albumin in Cardiac Surgery (ALBICS) trial was to compare the safety and effectiveness of 4% albumin (intervention) and Ringer acetate (conventional group) as the priming solution of cardiopulmonary bypass and intraoperative and postoperative volume replacement in patients undergoing on-pump cardiac surgery. The hypothesis was that compared with Ringer acetate, the albumin solution would reduce the number of patients with at least 1 major adverse event.

Methods

Trial Design

This trial was a single-center, randomized, double-blind, parallel-group trial to compare 4% albumin and Ringer acetate solutions in 1386 patients (in a 1:1 ratio) undergoing cardiac surgery with cardiopulmonary bypass at Helsinki University Hospital, Helsinki, Finland. We previously published the protocol (Supplement 1) and the statistical analysis plan (Supplement 2).⁸ Both the ethics committee of the Hospital District of Helsinki and Uusimaa, Helsinki, Finland, and the Finnish Medicines Agency approved the trial. All patients gave their written informed consent before participation. An independent data and safety monitoring board reviewed the planned interim analysis of the first 600 patients.

Patient Population

We prospectively screened all patients scheduled for a cardiac surgical procedure (Figure 1). Eligible patients were aged 18 through 90 years undergoing the following primary or repeat open heart surgery procedures (elective surgery or surgery during the index admission) independently or in combination: coronary artery bypass graft surgery; aortic, mitral, or tricuspid valve replacement or repair; aortic root or ascending aorta surgery without hypothermic circulatory arrest; or the maze procedure.

Exclusion criteria were immediate emergency surgery; congenital cardiac surgery; infection anticipated to compromise postprocedural rehabilitation; ongoing heart failure or low output syndrome (predefined significant inotropic support, mechanical ventilation, extracorporeal membrane oxygenation support, intra-aortic balloon pump, mechanical assistance of the left ventricle, left ventricular ejection fraction of less than 20%, or other comparable preoperative conditions); end-stage kidney disease (estimated glomerular filtration rate [GFR] <20 mL/min); hemophilia A or B; patient refusal of blood products or derivatives; and ticagrelor, prasugrel, clopidogrel, apixaban, or rivaroxaban use within 2 preoperative days or dabigatran use within 3 preoperative days.

Randomization

The HUS Pharmacy (the hospital pharmacy) independently performed randomization with computer-generated randomization lists using online software (http://www.randomization.com, which no longer offers off-line guidance) in a 1:1 ratio using variable block sizes (15 to 30). The patients, the entire study group, study nurses, and personnel taking care of the patients were blinded to randomization. For each patient, HUS Pharmacy delivered a study solution bag set with a unique allocation number.

Intervention

The trial consisted of 2 phases with the same study solution (4% albumin or Ringer acetate) in a double-blind fashion. First, study solutions were used for cardiopulmonary bypass priming. Second, during surgery and for the first 24 hours in the intensive care unit (ICU) or until discharged from the ICU (whichever occurred first), study solutions were given for volume replacement up to 3200 mL. If needed, Ringer acetate solution was used thereafter. Dosing of the study fluid was based on clinical decision. Apart from this intervention, the study had a pragmatic design.

Blinding

The HUS Pharmacy produced blinded study solution bags. Albuman 40 g/L, Albuman 200 g/L (Prothya Biosolutions) and Ringer acetate (Ringer Acetat Baxter Viaflo, Baxter) were packed in identical ethylene vinyl acetate bags (EVA Parenteral Nutrition Container, Baxter). The bags were placed in opaque cover bags (Opaque Protection Bag for Light Sensitive Drugs, Maco Pharma) labeled with a patient-specific allocation number.

In the first phase, the cardiopulmonary bypass circuit was initially primed with 1200 mL of Ringer acetate solution. After heparinization, blood was drawn to the cardiopulmonary bypass reservoir to mask the initial priming fluid color. The blinded study solution bag of 300 mL containing either 20% albumin solution or Ringer acetate solution was added using a nontransparent tube. The total priming volume (1500 mL) contained either the albumin solution at a final concentration of 4% or the pure Ringer acetate solution.

In the second phase, the opaque study solution bags were administered by infusion pumps using transparent, orange-colored infusion tubes (Infusomat, B Braun) with the drop chamber covered.

Outcomes

The primary effectiveness end point was the number of patients with at least 1 major adverse event (MAE) during the study period of 90 days. The composite end point of MAEs comprised (1) all-cause death; (2) myocardial injury (creatinine kinase muscle-brain isoenzymes, CK-MB, at least 10 times the upper reference limit, ie, 70 μg/L on the first postoperative morning); (3) new-onset of acute heart failure or low output syndrome requiring major intravenous inotropic support (the average dose during the intervention period: epinephrine ≥0.05 μg/kg/min, milrinone ≥0.5 μg/kg/min, or dobutamine ≥10 μg/kg/min), intra-aortic balloon pump support, extracorporeal membrane oxygenation support, or hospital readmission due to heart failure; (4) resternotomy including subxiphoidal pericardial drainage or other comparable procedure for any indication; (5) stroke; (6) major arrhythmia (ventricular fibrillation or ventricular tachycardia while off cardiopulmonary bypass, new-onset atrial fibrillation requiring permanent anticoagulation at the end on the 90-day follow-up, a new need for a permanent pacing device); (7) major bleeding (chest tube drainage at 18 hours after surgery of >20 mL/kg, need for ≥5 units of packed red blood cells, or an equivalent volume of washed red blood cells within the intervention period); (8) infection compromising postprocedural rehabilitation (a verification or strong suspicion of a systemic infection and/or an infection resulting in hospital readmission; reviewed by the authors E.P., H.V., and R.S.); and (9) acute kidney injury (postoperative creatinine level at least doubled from the preoperative level or kidney replacement therapy).

The secondary outcome measures were as follows: (1) total number of MAEs; (2) incidence of major adverse cardiac event (cardiac death, myocardial injury, acute heart failure, arrhythmia); (3) amount of each type of blood product transfused (red blood cells, fresh frozen plasma, platelets); (4) total fluid balance; (5) total measured blood loss (chest tube drainage); (6) acute kidney injury development; (7) days alive without mechanical ventilation during 90-day follow-up; (8) days alive outside ICU during 90-day follow-up; (9) days alive at home during 90-day follow-up; and (10) 90-day mortality.

Sample Size Calculation

Based on a previous cohort in this study site, we estimated the incidence of the primary end point to be 30%.⁹ Detection of an absolute difference of 7.5% (corresponding to a relative difference of 25%) between the study groups in the primary end point with a 2-sided P value of .05 and power of 80% required 1242 patients. In an analysis among the first 550 patients, the incidence of the primary end point was 42%. Due to increased incidence to retain the detection of the absolute difference of 7.5% between the groups, the detected relative difference decreased to 17.9%. Therefore, an independent statistician adapted the sample size to 1386 patients.

Statistical Analysis

Excluding statistical analyses indicated as post hoc analyses, we conducted all analyses in accordance with the predefined analysis plan (Supplement 2). Patients were analyzed according to their randomization group. After randomization but before initiating cardiopulmonary bypass, we excluded patients who, due to a major intraoperative change of the surgical plan, could not receive the study intervention or had fulfilled an exclusion criterion (hypothermic circulatory arrest). There were no missing primary end point data. We did not make any assumptions about missing data. We compared the primary outcome between the study groups using the Fisher exact test. In addition, a binomial regression model was used to estimate both relative risk (RR) and absolute differences in proportions with the primary end point and corresponding 95% CIs. An external data and safety monitoring board conducted an interim analysis after the first 600 patients. We employed the Lan-DeMets spending function with O’Brien-Fleming–type boundaries to preserve the overall 2-sided type I error rate at the .05 significance level. For the primary end point, the exact boundary values were P = .00132 in the interim analysis and P = .04868 in the final analysis.

For secondary outcomes, we compared the proportions with the Fisher exact test and binomial regression models. To evaluate differences between the treatment groups, we used the t test and linear regression models. We conducted prespecified subgroup analyses according to preoperative GFR (higher vs lower than the median value), EuroSCORE II (European System for Cardiac Operative Risk Evaluation; higher vs lower than the median value), and presence or absence of aortic stenosis as an indication (independently or in combination) for cardiac surgery. To evaluate whether the effect of intervention was the same in these subgroups, an interaction term between the indicator variable for the treatment, and each of the subgroup variables, respectively, was evaluated with the binomial regression model on the primary end point.

We conducted the safety analysis based on comparison of serious adverse events (SAEs) between the study groups. An MAE was always regarded as an SAE. Consequently, the final number of SAEs was the sum of MAEs and SAEs. We compared the study groups with 3 different statistical analyses: (1) the number of patients with at least 1 SAE (Fisher exact test and binomial regression models), (2) the SAE number per patient within those study patients with at least 1 SAE (t test), and (3) the SAE number per patient within all study patients (t test).

We conducted secondary outcome, subgroup, and safety analyses with a 2-sided significance level of .05 without predefined calculations of statistical power and without corrections for multiple testing; thus, these statistical analyses should be considered exploratory by nature. For post hoc analyses not included in the predefined statistical analysis plan, we tested interrelationships between single MAEs with the Fisher exact test and binomial regression models. We report binary outcome measures as frequencies (percentages) per treatment group and corresponding RRs with 95% CIs, continuous outcome measures as means with 95% CIs (symmetric distributions) or medians and interquartile ranges (IQRs) (skewed distributions), and treatment group differences as differences of means with 95% CIs. Stata software, version 15.1 (StataCorp), was used for statistical analyses.

Results

Patients

During the patient recruitment from March 21, 2017, to January 14, 2020 (last follow-up, April 13, 2020), 3627 patients underwent a cardiac surgical procedure. We excluded 1084 patients for procedure-related reasons and 624 patients for patient-related reasons (Figure 1). Of 1919 eligible patients, 1407 entered the study. We had to withdraw 21 patients after the start of surgery but before the onset of cardiopulmonary bypass due to changes of surgical plans making study intervention impossible (3 inoperable; 13 off pump) or for fulfilling an exclusion criterion (5 hypothermic circulatory arrest; Figure 1). Two patients fulfilling an exclusion criterion were accidentally enrolled. There were 33 patients (14 in the albumin group and 19 in the Ringer group) with protocol violations in fluid treatment (eTable 21 in Supplement 3). Baseline patient characteristics, preoperative laboratory values, and characteristics of surgical procedure were comparable between the study groups (Table 1).

Table 1. Patient, Operative, and Treatment Characteristics.

	Albumin (n = 693)	Ringer (n = 693)
Age, mean (SD), y	65 (10)	65 (10)
No. of patients, No. (%)
Men	549 (79)	542 (78)
Women	144 (21)	151 (22)
BMI, mean (SD)	28 (5)	28 (5)
EuroSCORE II, median (IQR)^a	1.7 (1.0-2.9)	1.7 (1.0-2.8)
NYHA class, No.^b
I	53	44
II	244	267
III	298	283
IV	98	99
Ejection fraction, %, No.
>50	537	579
31-50	140	104
≤30	16	10
Patients operated for aortic stenosis, No. (%)^c	147 (21)	160 (23)
Preoperative, mean (SD)
GFR, mL/min/1.73 m²^d	80 (17)	80 (17)
Hemoglobin, g/dL^e	14.1 (1.3)	14.1 (1.3)
Platelet count, ×10³/μL^f	230 (65)	230 (64)
INR, median (IQR)	1.0 (1.0-1.1)	1.0 (1.0-1.1)
Surgical procedures, No.
CABG surgery only	322	304
CABG + valve	50	59
1 valve only	161	159
2 valves only	8	11
Ascending aorta or aortic root only	22	29
≥2 Procedures^g	130	131
Cardiopulmonary bypass time, mean (SD), min	109 (40)	110 (44)
Aortic cross-clamp time, mean (SD), min	81 (33)	82 (34)
Administered study fluid, median (IQR), mL^h	2150 (1598-2700)	3298 (2669-3500)

Open in a new tab

Abbreviations: BMI, body mass index, calculated as weight in kilograms divided by height in meters squared; CABG, coronary artery bypass graft; GFR, glomerular filtration rate; INR, international normalized ratio; NYHA, New York Heart Association.

^{^a}

The European System for Cardiac Operative Risk Evaluation (EuroSCORE) II predicts in-hospital mortality risk in percentages, based on several surgical and physical criteria.

^{^b}

Range: class I, no physical activity limits, through class IV, discomfort from any physical activity.

^{^c}

Aortic stenosis alone or with other indications for surgery.

^{^d}

Reference values: 18 through 39 years, higher than 89 mL/min/1.73 m²; 40 through 49 years, higher than 83 mL/min/1.73 m²; 50 through 59 years, higher than 77 mL/min/1.73 m²; 60 through 69 years, higher than 69 mL/min/1.73 m²; and 70 years or older, higher than 59 mL/min/1.73 m².

^{^e}

Reference values: women, 11.7-15.5 g/dL; men, 13.4-16.7 g/dL.

^{^f}

Reference values: 150-360 ×10³/μL.

^{^g}

At least 2 procedures outside of other categories (ie, maze procedure, pulmonary vein isolation, closure of the atrial septal defect, patent foramen oval, patent ductus arteriosus, or left atrial appendage).

^{^h}

Only intravenous volume replacement. All patients received 300 mL of study fluid in cardiopulmonary bypass prime.

Trial Treatment

In addition to the cardiopulmonary bypass priming study fluid, patients in the albumin group received a median of 2150 mL (IQR, 1598-2700 mL) of the study fluid: 994 mL (IQR, 797-1292 mL) while in the operating room and 1033 mL (IQR, 643-1511 mL) while in the ICU. Patients in the Ringer group received a median of 3298 mL (IQR, 2669-3500 mL) of study fluid: 1909 mL (IQR, 1303-2700 mL) in the operating room and 919 mL (IQR, 460-1431 mL) in the ICU (P < .001).

Primary Outcome

At least 1 MAE occurred in 257 of 693 patients (37.1%) in the albumin group and in 234 of 693 patients (33.8%) in the Ringer group (Fisher exact test P = .22; RR of albumin/Ringer, 1.10; 95% CI, 0.95 to 1.27; absolute difference, 3.3 percentage points; 95% CI, −1.7 to 8.4; (Figure 2; eTables 1-9 in Supplement 3). The incidence of myocardial injury was significantly lower in the albumin group (RR, 0.44; 95% CI, 0.28 to 0.68; P < .001), but the incidences of bleeding (RR, 1.73; 95% CI, 1.12 to 2.68; P = .01), resternotomy (RR, 1.85; 95% CI, 1.28 to 2.68; P = .001), and infection (RR, 1.45; 95% CI, 1.07 to 1.97; P = .02) were significantly higher than in the Ringer group. No statistically significant differences existed in other components of the primary outcome measure (Figure 2).

Figure 2. — Major adverse events included all of the subsequent categories.

Secondary Outcomes

Total fluid balance during the intervention period was significantly lower in the albumin group (difference of albumin −Ringer, −1277 mL; 95% CI, −1433 to −1120 mL; P < .001). Chest tube drainage (difference, 178 mL; 95% CI, 138 to 219 mL; P < .001), as well as the administration of red blood cells (difference, 107 mL; 95% CI, 60 to 155 mL; P < .001) and platelets (difference, 71 mL; 95% CI, 41 to 100 mL; P < .001) were significantly higher in the albumin group than in the Ringer group (Table 2). No statistically significant differences existed in the other secondary outcome measures (Table 2).

Table 2. Secondary Outcomes.

	Albumin (n = 693)	Ringer (n = 693)	Albumin −Ringer, difference of means (95% CI)	P value
Intervention period
Fluid balance, mL			−1277 (−1433 to −1120)	<.001
Mean (95% CI)	4001 (3900 to 4102)	5278 (5158 to 5397)
Chest tube drainage, mL			178 (138 to 219)	<.001
Median (IQR)	720 (570 to 950)	590 (460 to 760)
Red blood cell transfusion, mL			107 (60 to 155)	<.001
Median (IQR)	0 (0 to 500)	0 (0 to 250)
Platelet transfusion, mL			71 (41 to 100)	<.001
Median (IQR)	0 (0 to 400)	0 (0 to 0)
Fresh frozen plasma transfusion, mL			−7 (−42 to 28)	.69
Median (IQR)	0 (0 to 0)	0 (0 to 0)
90-day Follow-up period
Days alive without ventilator			−0.1 (−0.6 to 0.4)	.64
Median (IQR)	89 (88 to 89)	89 (88 to 89)
Days alive outside ICU			−0.3 (−0.8 to 0.2)	.24
Median (IQR)	88 (88 to 88)	88 (88 to 88)
Days alive at home			−0.4 (−1.7 to 1.0)	.60
Median (IQR)	80 (75 to 82)	81 (76 to 82)
MACE, %^a	18.3	21.5	−3.2 (−7.4 to 1.0)	.14
No. of patients	127	149
Acute kidney injury, %^b	3.3	2.6	0.7 (−1.1 to 2.5)	.43
No of patients	23	18
Mortality, %	0.3	0.6	−0.3 (−1.0 to 0.4)	.42
No. of patients	2	4

Open in a new tab

Abbreviations: ICU, intensive care unit; MACE, major adverse cardiac event.

^{^a}

Composite outcome of cardiac death, myocardial injury, acute heart failure, and arrhythmia.

^{^b}

Postoperative creatinine level at least doubled from the preoperative level or kidney replacement therapy.

Subgroup Analyses

The incidence of infections in the patients with aortic stenosis and the fluid balance in the patients with high EuroSCORE were proportionally significantly greater in the albumin group. Otherwise, no differences existed in the predefined subgroups (eTables 12-20 in Supplement 3).

Post Hoc Analyses

The post hoc analysis found increased bleeding to be significantly associated with both resternotomy (RR, 11.16; 95% CI, 7.57 to 16.45; P < .001) and infection (RR, 3.16; 95% CI, 2.01 to 4.97; P < .001).

Safety

At least 1 SAE occurred in 103 of 693 patients (14.9%) in the Ringer group and in 114 of 693 patients (16.5%) in the albumin group (RR, 1.11; 95% CI, 0.87-1.41; P = .46). No significant differences were found in other indices of SAEs (eTables 10 and 11 in Supplement 3). The combined number of MAEs and SAEs was 283 of 693 (40.8%) in the Ringer group and 312 of 693 (45.0%) in the albumin group (RR, 1.10; 95% CI, 0.98 to 1.25; P = .13). The most common SAEs were pulmonary embolus (11 [1.6%] in the albumin group and 8 [1.2%] in the Ringer group), postpericardiotomy syndrome (9 [1.3%] in both groups), and pleural effusion with ICU or hospital readmission (7 [1.0%] in the albumin group and 9 [1.3%] in the Ringer group).

Discussion

Among patients undergoing cardiac surgery with cardiopulmonary bypass, treatment with 4% albumin solution for priming and perioperative intravenous volume replacement compared with Ringer acetate did not significantly reduce the risk of MAEs over the following 90 days. These findings do not support the use of 4% albumin solution in this setting.

Statistical analyses of the components of the composite primary outcome should be considered exploratory, given that the power of the trial was based on the composite outcome, as well as the potential for type I error due to multiple comparisons. However, single MAEs need to be analyzed to understand the result of the primary outcome. Albumin significantly reduced risk of myocardial injury but significantly increased risks of major bleeding, resternotomy, and infection. In previous studies, resternotomy and red blood cell transfusions were risk factors for infections.^10,11 In addition, high plasma albumin concentrations in the albumin group may have reduced the free plasma pool of antibiotics. To our knowledge, there have been only 2 RCTs comparing hemostasis between albumin and crystalloid in any surgical setting.^6,12 In a small RCT comprising on-pump cardiac surgical patients, postoperative chest tube drainage did not differ between patients receiving albumin (n = 76), hydroxyethyl starch (n = 81), or Ringer lactate (n = 79).⁶ However, compared with Ringer lactate, patients in the albumin group received red blood cells and fibrinogen concentrate more often and showed deteriorated thromboelastometry.⁶ In another RCT involving patients with radical cystectomy, no difference existed in blood loss between patients receiving albumin (n = 19) and Ringer lactate (n = 20).¹² However, albumin impaired hemostasis by thromboelastrography, activated partial thromboplastin time, and prothrombin time.¹² In an in vitro study excluding the effect of the endothelium, hypoalbuminemia enhanced hemostasis.¹³

In this study, albumin significantly reduced the risk of myocardial injury, defined as elevated postoperative plasma CK-MB in line with the cardioprotective properties of albumin.^14,15 Because blood cardioplegia was used, the cardioplegia solution was relatively rich in albumin in the albumin group. Addition of albumin to the storage solution conserved the coronary glycocalyx and improved heart function in experimental heart transplants in guinea pigs.¹⁵ However in this trial, incidences of acute heart failure and major arrhythmias did not significantly differ between the study groups, indicating that reduced myocardial biomarker concentration was not associated with reduced incidences of clinically important adverse cardiac events.

In previous studies, the effect of albumin on kidney function has been controversial.^6,7 In the present study, the acute kidney injury incidence (defined as Kidney Disease: Improving Global Outcomes stages 2 and 3 without urine output criteria) was only 3.0%, similar to the previously reported incidence of 3.3% in the study site.¹⁶ With such a low incidence, a significant difference was not detected in acute kidney injury between the study groups. In addition, the low mortality corroborates a previous finding from the study site.¹⁶ Due to low incidences, conclusions on the effects of albumin on acute kidney injury or mortality, even of an exploratory nature, cannot be drawn.

Limitations

This study has several limitations. First, it was a single-center trial in a tertiary university hospital, reducing generalizability of results. Second, most patients received either Ringer acetate or albumin solution alone for volume replacement for up to 24 postoperative hours. Thus, extrapolation of the study results to typical parallel clinical use of both crystalloids and colloids throughout the clinical course, especially if large volumes are needed, should be done with caution. Third, the 4% albumin and Ringer acetate used in this study differ from the widely used 5% albumin and Ringer lactate. Fourth, to interpret the primary outcome measure, patients already preoperatively fulfilling any of the primary outcome criteria or receiving adenosine diphosphate receptor inhibitors due to unstable coronary syndrome were excluded. This reduces generalizability of the trial findings to unstable patients.

Conclusions

Among patients undergoing cardiac surgery with cardiopulmonary bypass, treatment with 4% albumin for priming and perioperative intravenous volume replacement compared with Ringer acetate did not significantly reduce the risk of major adverse events over the following 90 days. These findings do not support the use of 4% albumin solution in this setting.

Supplement 1.

Trial Protocol

Click here for additional data file.^{(2.2MB, pdf)}

Supplement 2.

Statistical Analysis Plan

Click here for additional data file.^{(111.1KB, pdf)}

Supplement 3.

eTable 1. Death in the categorical time periods

eTable 2. Myocardial damage in the categorical time periods

eTable 3. Heart failure in the categorical time periods

eTable 4. Arrhythmia in the categorical time periods

eTable 5. Bleeding in the categorical time periods

eTable 6. Re-sternotomy in the categorical time periods

eTable 7. Stroke in the categorical time periods

eTable 8. Acute kidney injury in the categorical time periods

eTable 9. Infection in the categorical time periods

eTable 10. Events reported as serious adverse events (SAE)

eTable 11. The combined number of major adverse events (MAE) and events reported as serious adverse events (SAE)

eTable 12. Glomerular filtration rate (GFR) above or below the median (82.5 mL/min/1.73 m²)

eTable 13. Interaction of glomerular filtration rate (GFR) with the primary outcome measure and major adverse cardiac event (MACE)

eTable 14. Interaction of glomerular filtration rate (GFR) with the continuous secondary outcome measures

eTable 15. EuroSCORE II (European System For Cardiac Operative Risk Evaluation II) above or below the median (1.7)

eTable 16. Interaction of EuroSCORE II with the primary outcome measure and major adverse cardiac event (MACE)

eTable 17. Interaction of EuroSCORE II with the continuous secondary outcome measures

eTable 18. Patients with or without aortic stenosis

eTable 19. Interaction of the occurrence of aortic stenosis with the primary outcome measure and major adverse cardiac event (MACE)

eTable 20. Interaction of the occurrence of aortic stenosis with the secondary outcome measures

eTable 21. Protocol violations in fluid treatment: the amount of administered extra fluid

Click here for additional data file.^{(826KB, pdf)}

Supplement 4.

Data Sharing Statement

Click here for additional data file.^{(12.5KB, pdf)}

References

1.Verheij J, van Lingen A, Beishuizen A, et al. Cardiac response is greater for colloid than saline fluid loading after cardiac or vascular surgery. Intensive Care Med. 2006;32(7):1030-1038. doi: 10.1007/s00134-006-0195-5 [DOI] [PubMed] [Google Scholar]
2.Rex S, Scholz M, Weyland A, Busch T, Schorn B, Buhre W. Intra- and extravascular volume status in patients undergoing mitral valve replacement: crystalloid vs colloid priming of cardiopulmonary bypass. Eur J Anaesthesiol. 2006;23(1):1-9. doi: 10.1017/S0265021505001687 [DOI] [PubMed] [Google Scholar]
3.Zeng Y, Adamson RH, Curry FR, Tarbell JM. Sphingosine-1-phosphate protects endothelial glycocalyx by inhibiting syndecan-1 shedding. Am J Physiol Heart Circ Physiol. 2014;306(3):H363-H372. doi: 10.1152/ajpheart.00687.2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Belinskaia DA, Voronina PA, Shmurak VI, Jenkins RO, Goncharov NV. Serum albumin in health and disease: esterase, antioxidant, transporting and signaling properties. Int J Mol Sci. 2021;22(19):10318. doi: 10.3390/ijms221910318 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Russell JA, Navickis RJ, Wilkes MM. Albumin versus crystalloid for pump priming in cardiac surgery: meta-analysis of controlled trials. J Cardiothorac Vasc Anesth. 2004;18(4):429-437. doi: 10.1053/j.jvca.2004.05.019 [DOI] [PubMed] [Google Scholar]
6.Skhirtladze K, Base EM, Lassnigg A, et al. Comparison of the effects of albumin 5%, hydroxyethyl starch 130/0.4 6%, and Ringer’s lactate on blood loss and coagulation after cardiac surgery. Br J Anaesth. 2014;112(2):255-264. doi: 10.1093/bja/aet348 [DOI] [PubMed] [Google Scholar]
7.Lee EH, Kim WJ, Kim JY, et al. Effect of exogenous albumin on the incidence of postoperative acute kidney injury in patients undergoing off-pump coronary artery bypass surgery with a preoperative albumin level of less than 4.0 g/dL. Anesthesiology. 2016;124(5):1001-1011. doi: 10.1097/ALN.0000000000001051 [DOI] [PubMed] [Google Scholar]
8.Vlasov H, Juvonen T, Hiippala S, et al. Effect and safety of 4% albumin in the treatment of cardiac surgery patients: study protocol for the randomized, double-blind, clinical ALBICS (ALBumin In Cardiac Surgery) trial. Trials. 2020;21(1):235. doi: 10.1186/s13063-020-4160-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Petäjä L, Røsjø H, Mildh L, et al. Predictive value of high-sensitivity troponin T in addition to EuroSCORE II in cardiac surgery. Interact Cardiovasc Thorac Surg. 2016;23(1):133-141. doi: 10.1093/icvts/ivw060 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Phoon PHY, Hwang NC. Deep sternal wound infection: diagnosis, treatment and prevention. J Cardiothorac Vasc Anesth. 2020;34(6):1602-1613. doi: 10.1053/j.jvca.2019.09.019 [DOI] [PubMed] [Google Scholar]
11.Paternoster G, Guarracino F. Sepsis after cardiac surgery: from pathophysiology to management. J Cardiothorac Vasc Anesth. 2016;30(3):773-780. doi: 10.1053/j.jvca.2015.11.009 [DOI] [PubMed] [Google Scholar]
12.Rasmussen KC, Højskov M, Johansson PI, et al. Impact of albumin on coagulation competence and hemorrhage during major surgery: a randomized controlled trial. Medicine (Baltimore). 2016;95(9):e2720. doi: 10.1097/MD.0000000000002720 [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Paar M, Rossmann C, Nusshold C, et al. Anticoagulant action of low, physiologic, and high albumin levels in whole blood. PLoS One. 2017;12(8):e0182997. doi: 10.1371/journal.pone.0182997 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Mapanga RF, Joseph DE, Saieva M, et al. Glycation abolishes the cardioprotective effects of albumin during ex vivo ischemia-reperfusion. Physiol Rep. 2017;5(2):e13107. doi: 10.14814/phy2.13107 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Jacob M, Paul O, Mehringer L, et al. Albumin augmentation improves condition of guinea pig hearts after 4 hr of cold ischemia. Transplantation. 2009;87(7):956-965. doi: 10.1097/TP.0b013e31819c83b5 [DOI] [PubMed] [Google Scholar]
16.Petäjä L, Vaara S, Liuhanen S, et al. Acute kidney injury after cardiac surgery by complete KDIGO criteria predicts increased mortality. J Cardiothorac Vasc Anesth. 2017;31(3):827-836. doi: 10.1053/j.jvca.2016.08.026 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials