Intraoperative Oxygen Treatment, Oxidative Stress, and Organ Injury Following Cardiac Surgery: A Randomized Clinical Trial

Marcos G Lopez; Matthew S Shotwell; Cassandra Hennessy; Mias Pretorius; David R McIlroy; Melissa J Kimlinger; Eric H Mace; Tarek Absi; Ashish S Shah; Nancy J Brown; Frederic T Billings, IV

doi:10.1001/jamasurg.2024.2906

. 2024 Aug 7;159(10):1106–1116. doi: 10.1001/jamasurg.2024.2906

Intraoperative Oxygen Treatment, Oxidative Stress, and Organ Injury Following Cardiac Surgery

A Randomized Clinical Trial

Marcos G Lopez ¹, Matthew S Shotwell ², Cassandra Hennessy ², Mias Pretorius ¹, David R McIlroy ¹, Melissa J Kimlinger ⁶, Eric H Mace ³, Tarek Absi ⁴, Ashish S Shah ⁴, Nancy J Brown ⁵, Frederic T Billings IV ^1,^5,^✉, for the ROCS trial investigators

¹Department of Anesthesiology, Vanderbilt University Medical Center and Vanderbilt University School of Medicine, Nashville, Tennessee

²Department of Biostatistics, Vanderbilt University Medical Center and Vanderbilt University School of Medicine, Nashville, Tennessee

³Department of Surgery, Vanderbilt University Medical Center and Vanderbilt University School of Medicine, Nashville, Tennessee

⁴Department of Cardiac Surgery, Vanderbilt University Medical Center and Vanderbilt University School of Medicine, Nashville, Tennessee

⁵Department of Medicine, Vanderbilt University Medical Center and Vanderbilt University School of Medicine, Nashville, Tennessee

⁶Vanderbilt University Medical Center and Vanderbilt University School of Medicine, Nashville, Tennessee

Group Information: The ROCS trial investigators appear in Supplement 3.

Accepted for Publication: May 11, 2024.

Published Online: August 7, 2024. doi:10.1001/jamasurg.2024.2906

^✉

Corresponding Author: Frederic T. Billings IV, MD, MSc, Department of Anesthesiology, Vanderbilt University Medical Center and Vanderbilt University School of Medicine, Nashville, Tennessee, 1161 21st Ave S, T-4202 Medical Center North, Nashville, TN 37232 (frederic.t.billings@vumc.org).

Author Contributions: Drs Billings and Shotwell had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Shotwell, Hennessy, Brown, Billings.

Acquisition, analysis, or interpretation of data: Lopez, Shotwell, Hennessy, Pretorius, McIlroy, Kimlinger, Mace, Absi, Shah, Billings.

Drafting of the manuscript: Lopez, Shotwell, Hennessy, Mace, Absi, Billings.

Critical review of the manuscript for important intellectual content: Lopez, Shotwell, Hennessy, Pretorius, McIlroy, Kimlinger, Mace, Shah, Brown, Billings.

Statistical analysis: Lopez, Shotwell, Hennessy, Mace.

Obtained funding: Billings.

Administrative, technical, or material support: Lopez, Kimlinger, Mace, Brown.

Supervision: Lopez.

Conflict of Interest Disclosures: Dr Brown reported serving as science advisory board member for Alnylam Pharmaceuticals and eStar Biotech outside the submitted work. Dr Billings reported grants from the National Institutes of Health outside the submitted work. No other disclosures were reported.

Funding/Support: Dr Lopez received funding from the National Institutes of Health (NIH; K23GM129662). Dr Mace received funding from the NIH (T32GM108554). Dr Billings received funding from the NIH (R01GM112871 and R35GM145375). This study was also supported by NIH grant UL1TR000445 and the Vanderbilt University Medical Center Department of Anesthesiology.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Group Information: The ROCS trial investigators appear in Supplement 3.

Data Sharing Statement: See Supplement 4.

Additional Contributions: The authors acknowledge ROCS trial investigators from the Department of Anesthesiology at Vanderbilt University Medical Center: Robert Deegan, MD; Susan S. Eagle, MD; Antonio Hernandez, MD; Brian J. Gelfand, MD; Miklos D. Kertai, MD; Meredith Kingeter, MD; Ryan LeFevre, MD; Frederic W. Lombard, MD; Michael F. Mantinan, MD; Kelly K. Mishra, MD; Andrew D. Shaw, MB; Kara K. Siegrist, MD; and Ban Sileshi, MD, for contributions to protocol development, participant study, treatment administration, and final manuscript review and Robert E. Freundlich, MD, and Jonathan P. Wanderer, MD, for contributions to data acquisition and final manuscript review. We acknowledge the ROCS trial data and safety monitoring committee: Brian Donahue, MD, PhD (chair, Department of Anesthesiology, Vanderbilt University Medical Center); Gordon Bernard, MD (Department of Medicine, Vanderbilt University Medical Center); Benjamin Barton, MD (Department of Cardiac Surgery, Vanderbilt University Medical Center); and Daniel Byrne, MS (Department of Biostatistics, Vanderbilt University Medical Center). In addition, we acknowledge Anthony DeMatteo, BS (Vanderbilt University Medical Center), and Jing Zhou, PhD (Vanderbilt University Medical Center) for technical assistance; Stephanie Sanchez, BS (Vanderbilt University Medical Center), Ginger Milne, PhD (Vanderbilt University Medical Center), and the Eicosanoid Core (Vanderbilt University Medical Center) for assistance with quantifying F2-isoprostanes and isofurans; Jennifer L. Morse, MS (Department of Biostatistics, Vanderbilt University Medical Center), for data management assistance; Nicole Zaleski, MA, MPH (Vanderbilt University Medical Center), for assistance building the supplement and reviewing the manuscript; and Martha Tanner, BA (Department of Anesthesiology, Vanderbilt University Medical Center), for assistance preparing and uploading the manuscript. We acknowledge Tracie Baker, BA; Derwin Campbell, BA; Amanda Tilley, RN; Patricia Hendricks, RN; and Debra Craven, MS, of the of the Vanderbilt Department of Anesthesiology’s Perioperative Clinical Research Institute for assisting with participant study and data collection. We acknowledge Vanderbilt Department of Cardiac Surgery perfusionists Dane A. Fornero, BA; Phyllis Aycock, BA; Joseph Bianchi, BA; Joey Lepore, BA; Mark Meholchik, BA; Marina Mailyan, BA; Harry Moneypenny, BA; and Matthew Warhoover, BA, for contributions to participant study and treatment administration. We acknowledge Vanderbilt Department of Anesthesiology intensivists Bret Alvis, MD; Nathan Ashby, MD; Christina Boncyk, MD; Christina Hayhurst, MD; Christopher Patrick Henson, DO; Christina Jelly, MD; Phil Leisy, MD; Dorothee Mueller, MD; and Kimberly Rengel, MD, for contributions to postoperative participant management and outcome collection. No one acknowledged received compensation for their work, and all provided consent to be included. We thank the patients who agreed to participate in this study.

^✉

Corresponding author.

PMCID: PMC11307166 PMID: 39110454

Key Points

Question

Does administration of high concentrations of oxygen (hyperoxia) increase oxidative stress and organ injury compared to a strategy to maintain normal blood oxygen concentrations (normoxia) during cardiac surgery?

Findings

In this randomized clinical trial of 200 adult patients undergoing cardiac surgery, intraoperative hyperoxia compared to normoxia increased oxidative stress during surgery but did not affect postoperative kidney injury or other measurements of morbidity.

Meaning

Strategies to maximize or minimize patient oxygenation during cardiac surgery may not affect postoperative organ injury.

This randomized clinical trial evaluates the use of intraoperative hyperoxia vs normoxia in oxidative stress and organ injury following cardiac surgery.

Abstract

Importance

Liberal oxygen (hyperoxia) is commonly administered to patients during surgery, and oxygenation is known to impact mechanisms of perioperative organ injury.

Objective

To evaluate the effect of intraoperative hyperoxia compared to maintaining normoxia on oxidative stress, kidney injury, and other organ dysfunctions after cardiac surgery.

Design, Setting, and Participants

This was a participant- and assessor-blinded, randomized clinical trial conducted from April 2016 to October 2020 with 1 year of follow-up at a single tertiary care medical center. Adult patients (>18 years) presenting for elective open cardiac surgery without preoperative oxygen requirement, acute coronary syndrome, carotid stenosis, or dialysis were included. Of 3919 patients assessed, 2501 were considered eligible and 213 provided consent. Of these, 12 were excluded prior to randomization and 1 following randomization whose surgery was cancelled, leaving 100 participants in each group.

Interventions

Participants were randomly assigned to hyperoxia (1.00 fraction of inspired oxygen [FiO₂]) or normoxia (minimum FiO₂ to maintain oxygen saturation 95%-97%) throughout surgery.

Main Outcomes and Measures

Participants were assessed for oxidative stress by measuring F₂-isoprostanes and isofurans, for acute kidney injury (AKI), and for delirium, myocardial injury, atrial fibrillation, and additional secondary outcomes. Participants were monitored for 1 year following surgery.

Results

Two hundred participants were studied (median [IQR] age, 66 [59-72] years; 140 male and 60 female; 82 [41.0%] with diabetes). F₂-isoprostanes and isofurans (primary mechanistic end point) increased on average throughout surgery, from a median (IQR) of 73.3 (53.1-101.1) pg/mL at baseline to a peak of 85.5 (64.0-109.8) pg/mL at admission to the intensive care unit and were 9.2 pg/mL (95% CI, 1.0-17.4; P = .03) higher during surgery in patients assigned to hyperoxia. Median (IQR) change in serum creatinine (primary clinical end point) from baseline to postoperative day 2 was 0.01 mg/dL (−0.12 to 0.19) in participants assigned hyperoxia and −0.01 mg/dL (−0.16 to 0.19) in those assigned normoxia (median difference, 0.03; 95% CI, −0.04 to 0.10; P = .45). AKI occurred in 21 participants (21%) in each group. Intraoperative oxygen treatment did not affect additional acute organ injuries, safety events, or kidney, neuropsychological, and functional outcomes at 1 year.

Conclusions

Among adults receiving cardiac surgery, intraoperative hyperoxia increased intraoperative oxidative stress compared to normoxia but did not affect kidney injury or additional measurements of organ injury including delirium, myocardial injury, and atrial fibrillation.

Trial Registration

ClinicalTrials.gov Identifier: NCT02361944

Introduction

Approximately 2 million people receive cardiac surgery each year worldwide, and acute kidney injury affects 22% of these patients.¹ Brain dysfunction and injury, manifested as postoperative delirium, affect 25% of cardiac surgery patients,² and new-onset atrial fibrillation affects 35%.³ These acute organ injuries prolong hospitalization; increase long-term morbidity, such as chronic kidney disease, cognitive decline, and heart failure^4,5; and increase odds of death at 30 days by 500%.⁶ Ischemia reperfusion and oxidative injury are major contributors to postoperative organ injury,^7,8,9,10,11 and oxygen is integral to these molecular processes. It is unclear, however, if increasing oxygen tension during surgery by the administration of oxygen in excess of that required to saturate hemoglobin (hyperoxia) attenuates or exacerbates perioperative organ injury. Hyperoxia increases oxygen content in tissues,¹² may decrease regional hypoxia,¹³ and might limit poor clinical outcomes associated with perioperative ischemia.¹⁴ Conversely, hyperoxia may increase reactive oxygen species production,^15,16 may exacerbate reperfusion injury,^17,18 and has been associated with postoperative organ injury.¹⁹ We performed the Risk of Oxygen during Cardiac Surgery (ROCS) trial to test the hypothesis that intraoperative hyperoxia compared to a strategy to maintain normoxia increases oxidative stress, kidney injury, additional organ injuries, and postoperative morbidity.

Methods

Trial Design and Oversight

ROCS was an investigator-initiated, parallel-group, randomized clinical trial of hyperoxygenation (hyperoxia) vs intraoperative physiologic oxygenation (normoxia). The protocol and statistical analysis plan (Supplement 1) were detailed in National Institutes of Health grant R01GM112871, registered with ClinicalTrials.gov, published at the time of trial initiation,²⁰ and approved by the Vanderbilt University Medical Center institutional review board and an independent data and safety monitoring committee. The data and safety monitoring committee recommended continuing the study without changes to the protocol at 2 planned interim analyses that evaluated safety (not efficacy) of the intervention. All participants provided written informed consent. The study follows the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

Participants

Adult patients 18 years and older presenting for elective open cardiac surgery at Vanderbilt University Medical Center were eligible. Patients with acute coronary syndrome within 72 hours of surgery, home supplemental oxygen use, preoperative supplemental oxygen requirement, right to left intracardiac shunt, carotid stenosis greater than 50%, planned use of intraoperative circulatory arrest, current dialysis, or pregnancy were ineligible.

Randomization and Masking

Participants were randomly assigned to receive hyperoxia or normoxia using a randomization schedule accessed online immediately prior to induction of general anesthesia. Randomization used permuted blocks of random size 2 or 4, stratified by chronic kidney disease, defined as an estimated glomerular filtration rate less than 60 mL/min/1.73 m², and planned use of cardiopulmonary bypass. Participants, surgeons, intensivists, investigators assessing outcomes, and statisticians were blinded to treatment assignment. Clinical anesthesiologists and perfusionists were informed of treatment after randomization by the in-room study coordinator so that they could administer treatment.

Treatment

All participants received 1.00 fraction of inspired oxygen (FiO₂) during anesthesia induction and tracheal intubation. In participants assigned hyperoxia, the anesthesiologist administered 1.00 FiO₂ throughout mechanical ventilation, and the perfusionist administered 0.80 fraction of delivered oxygen (FdO₂) during cardiopulmonary bypass (eFigure 2 in Supplement 2). In participants assigned normoxia, the anesthesiologist reduced the FiO₂ to 0.21 after intubation and then titrated the FiO₂ to the lowest required to maintain arterial SpO₂ between 95% and 97%, but no less than 0.21; during cardiopulmonary bypass (CPB), the perfusionist titrated the FdO₂ using the inline co-oximeter to maintain a PaO₂ between 80 and 110 mm Hg (to convert to kPa, multiply by 0.133).

Treatment extended from the time of tracheal intubation to transfer to the intensive care unit (ICU) following surgery. All participants were ventilated with 1.00 FiO₂ during transfer and then 0.40 or higher at admission to the ICU to maintain SpO₂ greater than 95%. Subsequent postoperative oxygen administration and tracheal extubation were at the discretion of the intensivist.

Perioperative Participant Monitoring and Care

Anesthetic, surgical, and postoperative management were conducted according to institutional protocols and standard operating procedures, and blood was collected at 7 perioperative time points (eMethods 2-5 in Supplement 2). Each participant’s FiO₂, SpO₂, and PaO₂ during CPB, cerebral oxygenation and arterial, venous, and pulmonary blood pressure measurements were monitored continuously and cardiac output intermittently with pulse oximetry, in-line blood gas analysis, near infrared spectroscopy, an arterial line, and a pulmonary artery catheter, respectively, during surgery. Participants were assessed for outcomes and adverse events twice daily by research staff until hospital discharge and again at 12 months during a telephone interview, in addition to medical record review.

Outcomes and Safety Evaluation

The primary mechanistic end point was oxidative damage quantified as the sum of plasma concentrations of F₂-isoprostanes and isofurans collected during and after surgery, and the primary clinical end point was change in serum creatinine concentration from baseline to postoperative day 2. F₂-isoprostanes and isofurans are nonenzymatic products of arachidonic acid peroxidation, stable in biologic samples, and considered best-in-class markers for quantification of oxidative damage in vivo.^21,22 Secondary end points were additional markers of kidney injury (AKI), including Kidney Disease Improving Global Outcomes serum creatinine criteria²³; urinary markers neutrophil gelatinase–associated lipocalin (NGAL), tissue inhibitor of metalloproteinase-2 and insulinlike growth factor binding protein-7 (TIMP-2–IGFBP7), and dialysis^24,25,26; delirium incidence, duration, and severity assessed using the Confusion Assessment Method for the ICU (CAM-ICU) and the CAM-ICU-7 scale twice daily throughout participants’ ICU stay or for at least 3 days postoperatively; and plasma concentrations of the neuronal injury marker ubiquitin carboxyl-terminal hydrolase isozyme L1^9,27,28; myocardial injury, quantified by measuring serum myocardial b fraction of creatine kinase on postoperative day 1; atrial fibrillation diagnosed with continuous telemetry and electrocardiograms; arterial lactate; plasma cell–free hemoglobin; duration of mechanical ventilation; pneumonia; surgical site infection; time to first ICU discharge; and a composite outcome that included AKI, delirium, atrial fibrillation, myocardial infarction, stroke, transient ischemic attack, pneumonia, surgical site infection, or death. One-year follow-up data included assessments of kidney function, cognitive function (Short Blessed Test),²⁹ and activities of daily living. Safety events included intraoperative hypoxemia, defined as SpO₂ less than 90% or PaO₂ less than 70 mm Hg, myocardial infarction defined by a new Q wave on electrocardiogram, tracheal reintubation, transient ischemic attack, stroke, death, and adverse events.

Statistical Analyses

The primary analysis was implemented using a modified intention-to-treat method that included all participants who were randomized and received surgery, grouped according to treatment assignment at randomization, and was unadjusted for clinical factors. Sample size was based on data from the Statin AKI Cardiac Surgery randomized clinical trial and clinically significant treatment effects.^9,20,30 Two hundred participants provided 80% power to detect a mean (SD) 20 (50) pg/mL difference in F₂-isoprostanes and isofurans or a mean (SD) 0.15 (0.38) mg/dL difference in serum creatinine change between groups with a type I error rate of 5%. To accommodate sample attrition between enrollment and surgery, we anticipated recruiting approximately 220 patients to study 200.

The effects of hyperoxia vs normoxia on the mean plasma concentrations of F₂-isoprostanes and isofurans (primary mechanistic end point) were quantified using linear mixed effects regression methods, adjusting for the baseline measurement of F₂-isoprostanes and isofurans (modeled linearly) and the collection time point (modeled categorically), and allowing for interaction of treatment with time point. A random intercept, indexed by participant, was used to account for between-participant heterogeneity. A simultaneous Wald test was used to test for an overall treatment effect with respect to time (primary comparison) in the intraoperative period (ie, to test the hypothesis that the intervention effect is null across multiple time points, specifically 30 minutes into cardiopulmonary bypass or off-pump coronary artery grafting and following cessation of cardiopulmonary bypass or off-pump coronary grafting) and the postoperative period (ICU admission, 6 hours after ICU admission, postoperative day 1, and postoperative day 2). Treatment effects were summarized using an estimate and 95% confidence interval. A similar statistical approach was used to evaluate the effect of treatment on NGAL, TIMP-2–IGFBP7, ubiquitin carboxyl-terminal hydrolase isozyme L1, lactate, plasma-free hemoglobin, and delirium severity.

Continuous outcomes, including the change in serum creatinine from baseline to postoperative day 2 (primary clinical end point), were compared between treatment groups by estimating the median differences, and the associated P values were calculated using the Wald method.³¹ Binary outcomes were compared by estimating the risk difference, and the associated P value was calculated using the Agresti-Coull method.³² Ordinal outcomes were analyzed using proportional odds logistic regression.

Heterogeneity of treatment effects was examined across several factors known to be associated with increased oxidative stress or increased kidney injury, including age, sex, body mass index, current smoking status, diabetes, chronic kidney disease, coronary artery bypass surgery, combined surgery, CPB, and duration of surgery. For both the primary mechanistic and clinical end points, linear regression was used to test for an interaction between the subgroup defining variable and the treatment effect. Logistic regression was used to examine the relationships between intraoperative F₂-isoprostanes and isofurans and binary outcomes.

The trial examined several independent end points, including F₂-isoprostanes and isofurans and serum creatinine change. There were no overarching statistical hypotheses addressed that used multiple tests and that would warrant familywise type-I error control (ie, adjustment for multiple comparisons) across the independent hypotheses.^33,34 Thus, hypothesis testing was implemented to ensure a type-I error rate of 5% for each primary end point. Secondary end points were considered exploratory. Two-tailed P values less than .05 were considered significant. R version 3.4.1 (R Foundation) was used for all statistical analyses.

Results

From April 2016 to October 2019, 3919 patients were screened, and 1418 were excluded (Figure 1). Two hundred and thirteen patients provided consent (eFigure 5 in Supplement 2), 12 were excluded prior to randomization, and 201 were randomized. One participant’s surgery was subsequently cancelled, and this participant was excluded. Two hundred participants, 100 assigned to hyperoxia and 100 assigned to normoxia, were analyzed. The median (IQR) age of participants was 66 (59-72) years, 140 were male and 60 were female, and 82 (41.0%) had diabetes. The median (IQR) baseline estimated glomerular filtration rate was 71 (52-86) mL/min/1.73 m², the median (IQR) duration of surgery was 313 (267-369) minutes, and 167 participants (83.5%) had CPB during surgery. Baseline and procedural factors were similar between treatment groups (Table 1).

Table 1. Participant Characteristics.

Characteristic	No. (%)
Characteristic	Hyperoxia (n = 100)	Normoxia (n = 100)
Age, median (IQR), y	64 (60-72)	66 (58-71)
Sex
Male	66 (66)	74 (74)
Female	34 (34)	26 (26)
Race^a
African American	5 (5)	7 (7)
European American	92 (92)	92 (92)
Other^b	3 (3)	2 (2)
Medical history
Hypertension	84 (84)	77 (77)
Congestive heart failure	36 (36)	35 (35)
Left ventricular ejection fraction, median (IQR), %	55 (55-60)	55 (54-60)
Coronary artery disease	67 (67)	61 (61)
Myocardial infarction	20 (20)	21 (21)
Atrial fibrillation	23 (23)	26 (26)
Prior cardiac surgery	13 (13)	12 (12)
Peripheral vascular disease	11 (11)	14 (14)
Current smoking	15 (15)	10 (10)
Chronic obstructive pulmonary disease	6 (6)	9 (9)
Obstructive sleep apnea	21 (21)	14 (14)
Diabetes	38 (38)	44 (44)
Chronic kidney disease (stage III [eGFR<60 mL/min/1.73 m²] or worse)	42 (42)	43 (43)
Stroke	7 (7)	5 (5)
Dementia	0	2 (2)
Mini-Mental State Examination score, median (IQR)	29 (27-30)	29 (27-30)
Trails B score, median (IQR), s	105 (84-154)	103 (81-133)
Pain, median (IQR), numerical rating scale 0-10	3 (0-6)	3.5 (0-6)
Baseline vital signs
SpO₂, median (IQR), %	97 (96-99)	97 (96-99)
Heart rate, median (IQR), beats/min	68 (60-77)	71 (63-80)
Systolic blood pressure, median (IQR), mm Hg	132 (119-146)	126 (119-146)
Diastolic blood pressure, median (IQR), mm Hg	76 (65-84)	78 (69-83)
Cardiac index, median (IQR), L/min/m²^c	2.1 (1.8-2.4)	2.1 (1.8-2.4)
Central venous pressure, median (IQR), mm Hg^c	13 (10-19)	14 (12-17)
Body mass index, median (IQR)^d	30 (25-33)	29 (26-34)
Baseline laboratory data
PaO₂, median (IQR), mm Hg	76 (67-82)	77 (70-86)
Creatinine, median (IQR), mg/dL	1.08 (0.84-1.36)	1.04 (0.87-1.30)
Glomerular filtration, median (IQR), mL/min/1.73 m²^e	69 (50-83)	73 (53-87)
Hematocrit, median (IQR), %	40 (36-44)	42 (38-44)
Procedure characteristics
Coronary artery bypass grafting surgery	58 (58)	43 (43)
Valve surgery	58 (58)	53 (53)
Aorta surgery	2 (2)	10 (10)
Duration of surgery, median (IQR), min	314 (261-367)	307 (265-367)
Corticosteroid treatment, median (IQR)	6 (6)	7 (7)
Isoflurane, mean end tidal, median (IQR), %	0.61 (0.52-0.71)	0.58 (0.48-0.70)
Cardiopulmonary bypass use	83 (83)	83 (83)
Cardiopulmonary bypass time, median (IQR), min^f	110 (81-146)	116 (77-158)
Circulatory arrest	0	3 (3)
Circulatory arrest time, median (IQR), min^g	0 (0-0)	17 (15-27)

Open in a new tab

Abbreviations: eGFR, estimated glomerular filtration rate; PaO₂, partial pressure of oxygen in arterial blood; SpO₂, arterial hemoglobin oxygen saturation.

SI conversion factors: To convert creatinine to µmol/L, multiply by 76.25; hematocrit to proportion of 1.0, multiply by 0.01; PaO₂ to kPa, multiply by 0.133.

^{^a}

Race data were collected by direct participant query (open ended question) and reported as required by the study sponsor, the National Institutes of Health.

^{^b}

Other included Asian American, Native American, and other (did not identify with any of the groups listed). These groups were combined owing to small numbers.

^{^c}

Characteristic was assessed in the operating room prior to surgery but after initiation of randomized treatment.

^{^d}

Calculated as weight in kilograms divided by height in meters squared.

^{^e}

Estimated using the Chronic Kidney Disease Epidemiology Collaboration formula.

^{^f}

Among patients receiving cardiopulmonary bypass.

^{^g}

Among patients receiving circulatory arrest.

Intraoperative Oxygenation

The median (IQR) FiO₂ during ventilation and FdO₂ during CPB were 0.95 (0.94-0.95) and 0.80 (0.80-0.80), respectively, in participants assigned hyperoxia and 0.28 (0.24-0.35) and 0.50 (0.40-0.55) in participants assigned normoxia (Table 2, Figure 2). The median (IQR) SpO₂ during the case was 100% (100-100) in participants assigned hyperoxia and 97% (97-98) in participants assigned normoxia; the median PaO₂ was 377 mm Hg (332-426) in participants assigned hyperoxia and 99 mm Hg (96-105) in participants assigned normoxia. On average, the median (IQR) cerebral oximetry was 6.0% (0.0 to 14.4) above baseline throughout surgery in participants assigned hyperoxia and −7.2% (−14.4 to −0.2) in participants assigned normoxia.

Table 2. Intraoperative Oxygenation.

w	Median (IQR)		Median difference (95% CI)	P value
w	Hyperoxia	Normoxia	Median difference (95% CI)	P value
FiO₂ during ventilation, fraction	0.95 (0.94 to 0.95)	0.28 (0.24 to 0.35)	0.67 (0.65 to 0.68)	<.001
FdO₂ during cardiopulmonary bypass, fraction	0.80 (0.80 to 0.80)	0.50 (0.40 to 0.55)	0.30 (0.30 to 0.35)	<.001
SpO₂ during ventilation, %	100 (99 to 100)	97 (97 to 98)	3.0 (2.0 to 3.0)	<.001
SpO₂ <90% during ventilation, No. (%), incidence	24 (24)	75 (75)	−51 (−63 to −39)^a	<.001
SpO₂ <90% during ventilation, min	0	2.0 (1.0 to 4.9)	−1.9 (−2.0 to −1.0)	<.001
PaO₂ during cardiopulmonary bypass, mm Hg	377 (332 to 426)	100 (97 to 105)	276 (263 to 289)	<.001
PaO₂ <70 mm Hg during bypass, No. (%), incidence	0	61 (61)	−61 (−71 to −51)^a	<.001
PaO₂ <70 mm Hg during bypass, min	0	1.2 (0.3 to 3.9)	−1.2 (−1.5 to −0.8)	<.001
SvO₂ sampled from pulmonary artery, %	78 (71 to 83)	68 (65 to 73)	8 (5 to 10)	<.001
Cerebral oximetry relative to baseline, %	6.0 (0.0 to 14.4)	−7.2 (−14.4 to 0.2)	13.6 (10.3 to 16.9)	<.001
Cerebral oximetry <80% baseline, No. (%), incidence	55 (55)	86 (86)	−31 (−43 to −19)^a	<.001
Cerebral oximetry <80% baseline, min	0.3 (0.0 to 17.7)	27.2 (3.3 to 92.2)	−14.5 (−35.1 to −4.8)	<.001

Open in a new tab

Abbreviations: FdO₂, fraction of delivered oxygen; FiO₂, fraction of inspired oxygen; SpO₂, peripheral arterial hemoglobin oxygen saturation; PaO₂, partial pressure of oxygen in arterial blood; SvO₂, mixed venous hemoglobin oxygen saturation.

SI conversion factors: To convert PaO₂ to kPa, multiply by 0.133.

^{^a}

Absolute risk differences (95% CIs) shown.

Figure 2. — A, Fraction of inspired oxygen (FiO₂) and delivered oxygen (FdO₂) during mechanical ventilation and cardiopulmonary bypass (CPB). B, Arterial oxygen saturation (SpO₂) and mixed venous oxygen saturation (SvO₂). C, Arterial partial pressure of oxygen (PaO₂). D, Cerebral oximetry. Dark lines reflect the median values and shading the IQRs. Median (IQR) durations of time from entering the operating room to induction of anesthesia were 16 (14-21) minutes; starting CPB or off-pump coronary grafting, 121 (106-147) minutes; stopping CPB or off-pump coronary grafting, 245 (203-290) minutes; and leaving the operating room, 313 (267-369) minutes.

Outcomes

Oxidative stress (primary mechanistic end point), quantified as the sum of plasma F₂-isoprostanes and isofurans and measured at 7 perioperative time points, increased throughout surgery from a median (IQR) of 73.3 (53.1-101.1) pg/mL at baseline to a peak of 85.5 (64.0-109.8) pg/mL at admission to the ICU and was associated with increased postoperative organ injury. Participants with perioperative F₂-isoprostanes and isofurans at or above the 75th centile had 68.3% greater odds of AKI (95% CI, 11.7-253.5) and 89.3% greater odds of delirium (95% CI, 23.8-289.6), compared to those at or below the 25th centile (eTable 11 in Supplement 2).

Intraoperative hyperoxia, compared to normoxia, increased F₂-isoprostanes and isofurans during surgery (P = .02, overall intraoperative treatment effect assessing measurements from samples collected 30 minutes into CPB and following CPB, 2 degrees of freedom) (eFigure 6 in Supplement 2). For example, F₂-isoprostanes and isofurans were 9.2 pg/mL (95% CI, 1.0-17.4; P = .03) higher following CPB in participants treated with hyperoxia compared to in those treated with normoxia. This effect, however, did not extend into the postoperative period following termination of oxygen treatment (P = .30, overall postoperative treatment effect assessing measurements from samples collected at ICU admission, 6 hours after ICU admission, on postoperative day 1, and on postoperative day 2, 4 degrees of freedom).

Intraoperative oxygen treatment did not affect kidney injury. Baseline median (IQR) serum creatinine concentration was 1.08 (0.84-1.33) in participants assigned hyperoxia and 1.04 (0.88-1.28) mg/dL in participants assigned normoxia (to convert to µmol/L, multiply by 88.4). The median (IQR) change in serum creatinine (primary clinical end point) was 0.01 (−0.12 to 0.19) mg/dL from baseline to postoperative day 2 in participants assigned hyperoxia and −0.01 (−0.16 to 0.19) mg/dL in participants assigned normoxia (median difference, 0.03; 95% CI, −0.04 to 0.10; P = .45). AKI occurred in 21 participants (21%) in each group (absolute risk difference, 0%; 95% CI, −11 to 11). Rates of moderate and severe AKI and dialysis were not different between treatment groups (Figure 3), nor were perioperative urinary concentrations of NGAL or TIMP-2–IGFBP7.

Figure 3. — Data are presented as medians (IQRs) and absolute median differences (95% CIs) for continuous variables and as counts (percentages) and absolute risk differences (95% CIs) for categorical variables. P values were calculated using the Wald and Agresti-Coull methods for median differences and risk differences, respectively. In the Forest plot, we standardized the median difference for continuous variables such that 0.2 units on the risk difference scale represent the difference between the 75th and 25th centiles of the continuous variable. CK-MB indicates creatine kinase–myocardial band, measured the morning of postoperative day 1.

^aStage 2 or 3 acute kidney injury (AKI) based on Kidney Disease: Improving Global Outcomes consensus criteria.

^bAmong 38 participants who developed delirium.

^cMean confusion assessment method for the Confusion Assessment Method for the Intensive Care Unit 7 (CAM-ICU-7) score over the first 3 postoperative days.

^dAny of AKI, delirium, atrial fibrillation, myocardial infarction, stroke, transient ischemic attack, pneumonia, surgical site infection, or death.

In total, 22 participants (22%) developed delirium in the hyperoxia group and 16 participants (16%) in the normoxia group (absolute risk difference, 6.0%; 95% CI, −4.8 to 16.8). The duration of delirium was not different between treatment groups, nor the severity of delirium averaged over the first 3 postoperative days. Perioperative plasma concentrations of ubiquitin carboxyl-terminal hydrolase isozyme L1 were not affected by oxygen treatment.

Forty-six participants (46%) experienced postoperative atrial fibrillation in the hyperoxia group compared to 37 participants (37%) in the normoxia group (absolute risk difference, 9.0%; 95% CI, −4.6 to 22.6). Median (IQR) serum myocardial b fraction of creatine kinase on postoperative day 1 was 27.2 (11.5-58.3) ng/mL in participants assigned hyperoxia and 24.5 (13.4-40.8) ng/mL in participants assigned normoxia (median difference, 0.2 ng/mL; 95% CI, −5.8 to 7.3). Duration of postoperative mechanical ventilation, pneumonia, surgical site infection, ICU length of stay, and the composite outcome of organ injury and death were not different in participants assigned hyperoxia vs normoxia. Treatment effects on the primary mechanistic and clinical outcomes were not statistically different in subgroup analyses (eTables 12 and 13 in Supplement 2).

Safety End Points

Arterial hypoxemia was less common in participants assigned hyperoxia. Twenty-four participants (24%) assigned hyperoxia had SpO₂ less than 90% during ventilation for a median (IQR) duration of 3.0 (1.5-5.9) minutes, and no participants (0%) experienced a level of PaO₂ less than 70 mm Hg during CPB. In the group assigned normoxia, 75 participants (75%) had SpO₂ less than 90% during ventilation for a median (IQR) of 3.0 (1.0-5.2) minutes, and 61 participants (61%) had PaO₂ less than 70 mm Hg for a median (IQR) of 1.5 (0.8-4.5) minutes during CPB. However, at ICU admission, the PaO₂ was 18 mm Hg lower (95% CI, 7 to 30) in participants who had been treated with hyperoxia compared to in those who had been treated with normoxia. Median (IQR) arterial lactate, a marker of anaerobic metabolism, was 1.8 (1.3-2.9) mmol/L at ICU admission in the hyperoxia group and 1.5 (1.2-2.3) mmol/L in the normoxia group (median difference, 0.2; 95% CI, 0.0-0.4). Cerebral hypoxia, defined as cerebral oximetry less than 80% of baseline,⁸ was also less common during surgery among patients assigned hyperoxia. Participants assigned hyperoxia experienced a median (IQR) 0.3 (0.0-17.7) minutes of cerebral hypoxia compared to 27.2 (3.3-92.1) minutes in participants assigned normoxia.

Myocardial infarction, stroke, transient ischemic attacks, reintubation, and death were rare and not affected by oxygen treatment. There was some evidence that postoperative leukocytosis may be increased by hyperoxia. Additional prespecified adverse events were not different in participants assigned hyperoxia compared to in participants assigned normoxia (eFigure 7 in Supplement 2).

One-Year Follow-up

One year following surgery, 6 participants had died. Among the 194 living, 126 (64.9%) were available for 1-year follow-up, 69 in the hyperoxia group and 57 in the normoxia group. Zero participants assigned hyperoxia and 3 participants assigned normoxia were receiving kidney replacement therapy at 1 year. Among remaining participants, the median (IQR) estimated glomerular filtration rate was 60.4 (44.3-84.4) mL/min/1.73 m² in the hyperoxia group and 63.0 (50.8-84.8) 4 mL/min/1.73 m² in the normoxia group. One hundred nineteen of the 126 participants available for 1-year follow-up (91.5%) were able to complete cognitive assessments. Of these, 109 (91.6%) had normal to minimal impairment and 10 (8.4%) had minimal to moderate impairment using the Short Blessed Test, and there were no differences between oxygen treatment groups (eTable 15 in Supplement 2). Activities of daily living at 1 year were also not different between treatment groups (eTable 16 in Supplement 2).

Discussion

In this randomized clinical trial of oxygen therapy administered to adult patients during cardiac surgery, administration of hyperoxia treatment during surgery increased intraoperative oxidative stress but did not impact evidence of kidney, brain, or heart injury compared to maintaining normoxia. Prespecified safety events known to be associated with hypoxemia were also unaffected by intraoperative oxygen treatment.

Kidney, brain, or heart injury and dysfunction affect up to one-third of patients receiving cardiac surgery.^2,3,35 Putative mechanisms of these organ injuries include tissue ischemia and hypoxia, reperfusion, and oxidative stress.¹⁰ Oxygen tension affects these processes, but clinical trials to date have not demonstrated that hyperoxia or maintenance of normoxia during surgery improves outcomes. McGuiness and colleagues³⁶ randomly assigned 298 patients undergoing cardiac surgery to hyperoxia vs normoxia, finding no difference in the incidence or severity of AKI between treatment groups. However, normoxia was applied only during CPB, and the hyperoxic exposure that occurred before and following CPB may have eliminated any treatment effects. Furthermore, hyperoxia treatment was more moderate, with a mean PaO₂ of 178 mm Hg during CPB compared to 377 mm Hg in the current study. More recently, Nam and colleagues,³⁷ found in a cluster crossover trial that 0.8 FiO₂ during off-pump coronary artery bypass grafting surgery did not affect hospital length of stay vs 0.3 FiO₂ but did result in a 50% decrease in AKI. In addition, Shaefi et al³⁸ and Holse et al³⁹ reported results of randomized oxygen treatment trials targeting cognitive function and myocardial injury, respectively. These trials also had more modest use of liberal oxygen in the hyperoxia groups and targeted patients at lower risk compared to the current study. The trial by Holse et al³⁹ studied patients receiving noncardiac surgery, and the trial by Shaefi et al³⁸ limited study to those receiving coronary artery bypass graft surgery. Neither trial demonstrated a difference in clinical outcomes between oxygen treatment groups. The current study was designed so that the treatment groups represented oxygen administration at both ends of the spectrum of feasibility and was applied throughout the entire intraoperative period, maximizing the likelihood of identifying an effect of hyperoxia. Protocol adherence was assured by assigning a study coordinator to remain with participants and guide oxygen titration according to protocol. As a result, precision to hyperoxia and normoxia treatments was achieved in a population at high risk for organ injury.

In addition to clinical markers of injury, the current study targeted a mechanism of perioperative organ injury, oxidative stress. We did this to assess for evidence of a subclinical effect of treatment, and we used best available markers for quantifying oxidative damage in vivo. Indeed, perioperative F₂-isoprostanes and isofurans were increased in participants who developed AKI and delirium, and hyperoxia increased these markers during surgery, but intraoperative oxygen treatment did not affect postoperative concentrations of these markers, nor did intraoperative oxygen treatment affect organ injury. The FiO₂ was 1.00 during transport from the operating room to the ICU. Perhaps extending hyperoxia and normoxia oxygen treatments into the postoperative period, a time when many patients remain mechanically ventilated and are hemodynamically unstable, could lead to different results.

Limitations

This trial has several limitations. Many eligible patients were not approached due to size of the study team and the demanding nature of the intervention, sampling schedule, and follow-up. This could impact internal validity. We conducted a single-center trial to maximize control over experimental conditions in this proof-of-concept mechanistic trial, but this may impact generalizability and external validity. Oxygen treatment was not continued into the early postoperative period when susceptibility to oxidative stress and organ injury is likely ongoing. The size of the cohort also limited power to identify small but clinically important differences in secondary end points and clinical subgroups and could increase risk of type II error. For example, the differences in delirium and atrial fibrillation, if real, are clinically significant. Results from this study have implications for future research. A large multicenter efficacy trial could accurately define the impact of hyperoxia during cardiac surgery on organ injury, and oxygen treatments could be extended into the postoperative period. The safety profile of extreme hyperoxia in the current study supports the safety of this intervention for the investigation in additional patient populations.

Conclusions

In conclusion, among adults receiving elective cardiac surgery, intraoperative hyperoxia increased intraoperative oxidative stress but did not affect kidney injury or additional measurements of organ injury and function compared to a strategy to maintain intraoperative normoxia.

Supplement 1.

Trial protocol and statistical analysis plan

jamasurg-e242906-s001.pdf^{(2.3MB, pdf)}

Supplement 2.

eTable 1. Participating Investigators

eTable 2. DSMC Members

eFigure 1. Schedule of Events

eTable 3. Inclusion criteria

eTable 4. Exclusion criteria

eFigure 2. Schematic of oxygen titration protocol

eTable 5. Primary and Secondary Endpoints

eTable 6. 1-Year Follow-Up Outcomes

eTable 7. Safety Events

eFigure 3. The CAM-ICU-7 Delirium Severity Scale

eFigure 4. Short Blessed Test

eTable 8. Short Blessed Test error allowances and weighting

eFigure 5. Participant recruitment

eTable 9. Effects of Treatment on Mean Arterial Pressure and Cardiac Index in the perioperative period

eFigure 6. Biomarkers in participants assigned hyperoxia and normoxia

eTable 10. Effects of treatment on F2-isoprostanes and isofurans during the perioperative period

eTable 11. Associations between perioperative oxidative stress and organ injury outcomes

eTable 12. Heterogeneity of treatment effect on primary mechanistic outcome, plasma concentrations of F2-isoprostanes and isofurans

eTable 13. Heterogeneity of treatment effect on primary clinical outcome, change in serum creatinine from baseline to postoperative day 2

eFigure 7. Adverse event plots

eTable 14. Postoperative oxygen administration

eTable 15. Cognitive function at 1 year (Short Blessed Test)

eTable 16. Instrumental Activities of Daily Living (ADLs)

eReferences

jamasurg-e242906-s002.pdf^{(3.8MB, pdf)}

Supplement 3.

The ROCS trial investigators

jamasurg-e242906-s003.pdf^{(116.6KB, pdf)}

Supplement 4.

Data sharing statement

jamasurg-e242906-s004.pdf^{(14.2KB, pdf)}

References

1.Hu J, Chen R, Liu S, Yu X, Zou J, Ding X. Global incidence and outcomes of adult patients with acute kidney injury after cardiac surgery: a systematic review and meta-analysis. J Cardiothorac Vasc Anesth. 2016;30(1):82-89. doi: 10.1053/j.jvca.2015.06.017 [DOI] [PubMed] [Google Scholar]
2.Brown CH IV, Probert J, Healy R, et al. Cognitive decline after delirium in patients undergoing cardiac surgery. Anesthesiology. 2018;129(3):406-416. doi: 10.1097/ALN.0000000000002253 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Yadava M, Hughey AB, Crawford TC. Postoperative atrial fibrillation: incidence, mechanisms, and clinical correlates. Heart Fail Clin. 2016;12(2):299-308. doi: 10.1016/j.hfc.2015.08.023 [DOI] [PubMed] [Google Scholar]
4.Pandharipande PP, Girard TD, Jackson JC, et al. ; BRAIN-ICU Study Investigators . Long-term cognitive impairment after critical illness. N Engl J Med. 2013;369(14):1306-1316. doi: 10.1056/NEJMoa1301372 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.LaPar DJ, Speir AM, Crosby IK, et al. ; Investigators for the Virginia Cardiac Surgery Quality Initiative . Postoperative atrial fibrillation significantly increases mortality, hospital readmission, and hospital costs. Ann Thorac Surg. 2014;98(2):527-533. doi: 10.1016/j.athoracsur.2014.03.039 [DOI] [PubMed] [Google Scholar]
6.Loef BG, Epema AH, Smilde TD, et al. Immediate postoperative renal function deterioration in cardiac surgical patients predicts in-hospital mortality and long-term survival. J Am Soc Nephrol. 2005;16(1):195-200. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=15563558. doi: 10.1681/ASN.2003100875 [DOI] [PubMed] [Google Scholar]
7.Billings FT IV, Pretorius M, Schildcrout JS, et al. Obesity and oxidative stress predict AKI after cardiac surgery. J Am Soc Nephrol. 2012;23(7):1221-1228. doi: 10.1681/ASN.2011090940 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Lopez MG, Pandharipande P, Morse J, et al. Intraoperative cerebral oxygenation, oxidative injury, and delirium following cardiac surgery. Free Radic Biol Med. 2017;103:192-198. doi: 10.1016/j.freeradbiomed.2016.12.039 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Lopez MG, Hughes CG, DeMatteo A, et al. Intraoperative oxidative damage and delirium after cardiac surgery. Anesthesiology. 2020;132(3):551-561. doi: 10.1097/ALN.0000000000003016 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Eltzschig HK, Eckle T. Ischemia and reperfusion–from mechanism to translation. Nat Med. 2011;17(11):1391-1401. doi: 10.1038/nm.2507 [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Walsh M, Devereaux PJ, Garg AX, et al. Relationship between intraoperative mean arterial pressure and clinical outcomes after noncardiac surgery: toward an empirical definition of hypotension. Anesthesiology. 2013;119(3):507-515. doi: 10.1097/ALN.0b013e3182a10e26 [DOI] [PubMed] [Google Scholar]
12.Rosenthal G, Hemphill JC III, Sorani M, et al. Brain tissue oxygen tension is more indicative of oxygen diffusion than oxygen delivery and metabolism in patients with traumatic brain injury. Crit Care Med. 2008;36(6):1917-1924. doi: 10.1097/CCM.0b013e3181743d77 [DOI] [PubMed] [Google Scholar]
13.Nortje J, Coles JP, Timofeev I, et al. Effect of hyperoxia on regional oxygenation and metabolism after severe traumatic brain injury: preliminary findings. Crit Care Med. 2008;36(1):273-281. doi: 10.1097/01.CCM.0000292014.60835.15 [DOI] [PubMed] [Google Scholar]
14.Mattishent K, Thavarajah M, Sinha A, et al. Safety of 80% vs 30-35% fraction of inspired oxygen in patients undergoing surgery: a systematic review and meta-analysis. Br J Anaesth. 2019;122(3):311-324. doi: 10.1016/j.bja.2018.11.026 [DOI] [PubMed] [Google Scholar]
15.Hoffman DL, Brookes PS. Oxygen sensitivity of mitochondrial reactive oxygen species generation depends on metabolic conditions. J Biol Chem. 2009;284(24):16236-16245. doi: 10.1074/jbc.M809512200 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Mace EH, Kimlinger MJ, No TJ, et al. Soluble guanylyl cyclase activation rescues hyperoxia-induced dysfunction of vascular relaxation. Shock. 2022;58(4):280-286. doi: 10.1097/SHK.0000000000001982 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Reilly MP, Delanty N, Roy L, et al. Increased formation of the isoprostanes IPF2alpha-I and 8-epi-prostaglandin F2alpha in acute coronary angioplasty: evidence for oxidant stress during coronary reperfusion in humans. Circulation. 1997;96(10):3314-3320. doi: 10.1161/01.CIR.96.10.3314 [DOI] [PubMed] [Google Scholar]
18.Chouchani ET, Pell VR, Gaude E, et al. Ischaemic accumulation of succinate controls reperfusion injury through mitochondrial ROS. Nature. 2014;515(7527):431-435. doi: 10.1038/nature13909 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.McIlroy DR, Shotwell MS, Lopez MG, et al. ; Multicenter Perioperative Outcomes Group . Oxygen administration during surgery and postoperative organ injury: observational cohort study. BMJ. 2022;379:e070941. doi: 10.1136/bmj-2022-070941 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Lopez MG, Pretorius M, Shotwell MS, et al. The Risk of Oxygen During Cardiac Surgery (ROCS) trial: study protocol for a randomized clinical trial. Trials. 2017;18(1):295. doi: 10.1186/s13063-017-2021-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Kadiiska MB, Gladen BC, Baird DD, et al. Biomarkers of oxidative stress study II: are oxidation products of lipids, proteins, and DNA markers of CCl4 poisoning? Free Radic Biol Med. 2005;38(6):698-710. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=15721980. doi: 10.1016/j.freeradbiomed.2004.09.017 [DOI] [PubMed] [Google Scholar]
22.Milne GL, Sanchez SC, Musiek ES, Morrow JD. Quantification of F2-isoprostanes as a biomarker of oxidative stress. Nat Protoc. 2007;2(1):221-226. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=17401357. doi: 10.1038/nprot.2006.375 [DOI] [PubMed] [Google Scholar]
23.KDIGO AKI Work Group . KDIGO clinical practice guideline for acute kidney injury. Kidney Int Suppl. 2012;2:1-138. [Google Scholar]
24.Mishra J, Dent C, Tarabishi R, et al. Neutrophil gelatinase-associated lipocalin (NGAL) as a biomarker for acute renal injury after cardiac surgery. Lancet. 2005;365(9466):1231-1238. doi: 10.1016/S0140-6736(05)74811-X [DOI] [PubMed] [Google Scholar]
25.Cummings JJ, Shaw AD, Shi J, Lopez MG, O’Neal JB, Billings FT IV. Intraoperative prediction of cardiac surgery-associated acute kidney injury using urinary biomarkers of cell cycle arrest. J Thorac Cardiovasc Surg. 2019;157(4):1545-1553.e5. doi: 10.1016/j.jtcvs.2018.08.090 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Kashani K, Al-Khafaji A, Ardiles T, et al. Discovery and validation of cell cycle arrest biomarkers in human acute kidney injury. Crit Care. 2013;17(1):R25. doi: 10.1186/cc12503 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Ely EW, Inouye SK, Bernard GR, et al. Delirium in mechanically ventilated patients: validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU). JAMA. 2001;286(21):2703-2710. doi: 10.1001/jama.286.21.2703 [DOI] [PubMed] [Google Scholar]
28.Lindroth H, Khan BA, Carpenter JS, et al. Delirium severity trajectories and outcomes in ICU patients. defining a dynamic symptom phenotype. Ann Am Thorac Soc. 2020;17(9):1094-1103. doi: 10.1513/AnnalsATS.201910-764OC [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Katzman R, Brown T, Fuld P, Peck A, Schechter R, Schimmel H. Validation of a short Orientation-Memory-Concentration Test of cognitive impairment. Am J Psychiatry. 1983;140(6):734-739. doi: 10.1176/ajp.140.6.734 [DOI] [PubMed] [Google Scholar]
30.Billings FT IV, Hendricks PA, Schildcrout JS, et al. High-dose perioperative atorvastatin and acute kidney injury following cardiac surgery: a randomized clinical trial. JAMA. 2016;315(9):877-888. doi: 10.1001/jama.2016.0548 [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Bauer DF. Constructing confidence sets using rank statistics. J Am Stat Assoc. 1972;67(339):687-690. doi: 10.1080/01621459.1972.10481279 [DOI] [Google Scholar]
32.Fagerland MW, Lydersen S, Laake P. Recommended confidence intervals for two independent binomial proportions. Stat Methods Med Res. 2015;24(2):224-254. doi: 10.1177/0962280211415469 [DOI] [PubMed] [Google Scholar]
33.Althouse AD. Adjust for multiple comparisons? it’s not that simple. Ann Thorac Surg. 2016;101(5):1644-1645. doi: 10.1016/j.athoracsur.2015.11.024 [DOI] [PubMed] [Google Scholar]
34.Feise RJ. Do multiple outcome measures require P value adjustment? BMC Med Res Methodol. 2002;2:8. doi: 10.1186/1471-2288-2-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Rosner MH, Okusa MD. Acute kidney injury associated with cardiac surgery. Clin J Am Soc Nephrol. 2006;1(1):19-32. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=17699187. doi: 10.2215/CJN.00240605 [DOI] [PubMed] [Google Scholar]
36.McGuinness SP, Parke RL, Drummond K, et al. ; SO-COOL investigators . A multicenter, randomized, controlled phase IIB trial of avoidance of hyperoxemia during cardiopulmonary bypass. Anesthesiology. 2016;125(3):465-473. doi: 10.1097/ALN.0000000000001226 [DOI] [PubMed] [Google Scholar]
37.Nam K, Nam JS, Kim HB, et al. ; Cardiac Surgery and Oxygen Therapy (CARROT) Investigators . Effects of intraoperative inspired oxygen fraction (FiO₂ 0.3 vs 0.8) on patients undergoing off-pump coronary artery bypass grafting: the CARROT multicenter, cluster-randomized trial. Crit Care. 2023;27(1):286. doi: 10.1186/s13054-023-04558-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Shaefi S, Shankar P, Mueller AL, et al. Intraoperative oxygen concentration and neurocognition after cardiac surgery. Anesthesiology. 2021;134(2):189-201. doi: 10.1097/ALN.0000000000003650 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Holse C, Aasvang EK, Vester-Andersen M, et al. ; VIXIE Trial Group . Hyperoxia and antioxidants for myocardial injury in noncardiac surgery: a 2 × 2 factorial, blinded, randomized clinical trial. Anesthesiology. 2022;136(3):408-419. doi: 10.1097/ALN.0000000000004117 [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplement 1.

Trial protocol and statistical analysis plan

jamasurg-e242906-s001.pdf^{(2.3MB, pdf)}

Supplement 2.

eTable 1. Participating Investigators

eTable 2. DSMC Members

eFigure 1. Schedule of Events

eTable 3. Inclusion criteria

eTable 4. Exclusion criteria

eFigure 2. Schematic of oxygen titration protocol

eTable 5. Primary and Secondary Endpoints

eTable 6. 1-Year Follow-Up Outcomes

eTable 7. Safety Events

eFigure 3. The CAM-ICU-7 Delirium Severity Scale

eFigure 4. Short Blessed Test

eTable 8. Short Blessed Test error allowances and weighting

eFigure 5. Participant recruitment

eTable 9. Effects of Treatment on Mean Arterial Pressure and Cardiac Index in the perioperative period

eFigure 6. Biomarkers in participants assigned hyperoxia and normoxia

eTable 10. Effects of treatment on F2-isoprostanes and isofurans during the perioperative period

eTable 11. Associations between perioperative oxidative stress and organ injury outcomes

eTable 12. Heterogeneity of treatment effect on primary mechanistic outcome, plasma concentrations of F2-isoprostanes and isofurans

eTable 13. Heterogeneity of treatment effect on primary clinical outcome, change in serum creatinine from baseline to postoperative day 2

eFigure 7. Adverse event plots

eTable 14. Postoperative oxygen administration

eTable 15. Cognitive function at 1 year (Short Blessed Test)

eTable 16. Instrumental Activities of Daily Living (ADLs)

eReferences

jamasurg-e242906-s002.pdf^{(3.8MB, pdf)}

Supplement 3.

The ROCS trial investigators

jamasurg-e242906-s003.pdf^{(116.6KB, pdf)}

Supplement 4.

Data sharing statement

jamasurg-e242906-s004.pdf^{(14.2KB, pdf)}

[soi240053r1] 1.Hu J, Chen R, Liu S, Yu X, Zou J, Ding X. Global incidence and outcomes of adult patients with acute kidney injury after cardiac surgery: a systematic review and meta-analysis. J Cardiothorac Vasc Anesth. 2016;30(1):82-89. doi: 10.1053/j.jvca.2015.06.017 [DOI] [PubMed] [Google Scholar]

[soi240053r2] 2.Brown CH IV, Probert J, Healy R, et al. Cognitive decline after delirium in patients undergoing cardiac surgery. Anesthesiology. 2018;129(3):406-416. doi: 10.1097/ALN.0000000000002253 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r3] 3.Yadava M, Hughey AB, Crawford TC. Postoperative atrial fibrillation: incidence, mechanisms, and clinical correlates. Heart Fail Clin. 2016;12(2):299-308. doi: 10.1016/j.hfc.2015.08.023 [DOI] [PubMed] [Google Scholar]

[soi240053r4] 4.Pandharipande PP, Girard TD, Jackson JC, et al. ; BRAIN-ICU Study Investigators . Long-term cognitive impairment after critical illness. N Engl J Med. 2013;369(14):1306-1316. doi: 10.1056/NEJMoa1301372 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r5] 5.LaPar DJ, Speir AM, Crosby IK, et al. ; Investigators for the Virginia Cardiac Surgery Quality Initiative . Postoperative atrial fibrillation significantly increases mortality, hospital readmission, and hospital costs. Ann Thorac Surg. 2014;98(2):527-533. doi: 10.1016/j.athoracsur.2014.03.039 [DOI] [PubMed] [Google Scholar]

[soi240053r6] 6.Loef BG, Epema AH, Smilde TD, et al. Immediate postoperative renal function deterioration in cardiac surgical patients predicts in-hospital mortality and long-term survival. J Am Soc Nephrol. 2005;16(1):195-200. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=15563558. doi: 10.1681/ASN.2003100875 [DOI] [PubMed] [Google Scholar]

[soi240053r7] 7.Billings FT IV, Pretorius M, Schildcrout JS, et al. Obesity and oxidative stress predict AKI after cardiac surgery. J Am Soc Nephrol. 2012;23(7):1221-1228. doi: 10.1681/ASN.2011090940 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r8] 8.Lopez MG, Pandharipande P, Morse J, et al. Intraoperative cerebral oxygenation, oxidative injury, and delirium following cardiac surgery. Free Radic Biol Med. 2017;103:192-198. doi: 10.1016/j.freeradbiomed.2016.12.039 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r9] 9.Lopez MG, Hughes CG, DeMatteo A, et al. Intraoperative oxidative damage and delirium after cardiac surgery. Anesthesiology. 2020;132(3):551-561. doi: 10.1097/ALN.0000000000003016 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r10] 10.Eltzschig HK, Eckle T. Ischemia and reperfusion–from mechanism to translation. Nat Med. 2011;17(11):1391-1401. doi: 10.1038/nm.2507 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r11] 11.Walsh M, Devereaux PJ, Garg AX, et al. Relationship between intraoperative mean arterial pressure and clinical outcomes after noncardiac surgery: toward an empirical definition of hypotension. Anesthesiology. 2013;119(3):507-515. doi: 10.1097/ALN.0b013e3182a10e26 [DOI] [PubMed] [Google Scholar]

[soi240053r12] 12.Rosenthal G, Hemphill JC III, Sorani M, et al. Brain tissue oxygen tension is more indicative of oxygen diffusion than oxygen delivery and metabolism in patients with traumatic brain injury. Crit Care Med. 2008;36(6):1917-1924. doi: 10.1097/CCM.0b013e3181743d77 [DOI] [PubMed] [Google Scholar]

[soi240053r13] 13.Nortje J, Coles JP, Timofeev I, et al. Effect of hyperoxia on regional oxygenation and metabolism after severe traumatic brain injury: preliminary findings. Crit Care Med. 2008;36(1):273-281. doi: 10.1097/01.CCM.0000292014.60835.15 [DOI] [PubMed] [Google Scholar]

[soi240053r14] 14.Mattishent K, Thavarajah M, Sinha A, et al. Safety of 80% vs 30-35% fraction of inspired oxygen in patients undergoing surgery: a systematic review and meta-analysis. Br J Anaesth. 2019;122(3):311-324. doi: 10.1016/j.bja.2018.11.026 [DOI] [PubMed] [Google Scholar]

[soi240053r15] 15.Hoffman DL, Brookes PS. Oxygen sensitivity of mitochondrial reactive oxygen species generation depends on metabolic conditions. J Biol Chem. 2009;284(24):16236-16245. doi: 10.1074/jbc.M809512200 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r16] 16.Mace EH, Kimlinger MJ, No TJ, et al. Soluble guanylyl cyclase activation rescues hyperoxia-induced dysfunction of vascular relaxation. Shock. 2022;58(4):280-286. doi: 10.1097/SHK.0000000000001982 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r17] 17.Reilly MP, Delanty N, Roy L, et al. Increased formation of the isoprostanes IPF2alpha-I and 8-epi-prostaglandin F2alpha in acute coronary angioplasty: evidence for oxidant stress during coronary reperfusion in humans. Circulation. 1997;96(10):3314-3320. doi: 10.1161/01.CIR.96.10.3314 [DOI] [PubMed] [Google Scholar]

[soi240053r18] 18.Chouchani ET, Pell VR, Gaude E, et al. Ischaemic accumulation of succinate controls reperfusion injury through mitochondrial ROS. Nature. 2014;515(7527):431-435. doi: 10.1038/nature13909 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r19] 19.McIlroy DR, Shotwell MS, Lopez MG, et al. ; Multicenter Perioperative Outcomes Group . Oxygen administration during surgery and postoperative organ injury: observational cohort study. BMJ. 2022;379:e070941. doi: 10.1136/bmj-2022-070941 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r20] 20.Lopez MG, Pretorius M, Shotwell MS, et al. The Risk of Oxygen During Cardiac Surgery (ROCS) trial: study protocol for a randomized clinical trial. Trials. 2017;18(1):295. doi: 10.1186/s13063-017-2021-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r21] 21.Kadiiska MB, Gladen BC, Baird DD, et al. Biomarkers of oxidative stress study II: are oxidation products of lipids, proteins, and DNA markers of CCl4 poisoning? Free Radic Biol Med. 2005;38(6):698-710. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=15721980. doi: 10.1016/j.freeradbiomed.2004.09.017 [DOI] [PubMed] [Google Scholar]

[soi240053r22] 22.Milne GL, Sanchez SC, Musiek ES, Morrow JD. Quantification of F2-isoprostanes as a biomarker of oxidative stress. Nat Protoc. 2007;2(1):221-226. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=17401357. doi: 10.1038/nprot.2006.375 [DOI] [PubMed] [Google Scholar]

[soi240053r23] 23.KDIGO AKI Work Group . KDIGO clinical practice guideline for acute kidney injury. Kidney Int Suppl. 2012;2:1-138. [Google Scholar]

[soi240053r24] 24.Mishra J, Dent C, Tarabishi R, et al. Neutrophil gelatinase-associated lipocalin (NGAL) as a biomarker for acute renal injury after cardiac surgery. Lancet. 2005;365(9466):1231-1238. doi: 10.1016/S0140-6736(05)74811-X [DOI] [PubMed] [Google Scholar]

[soi240053r25] 25.Cummings JJ, Shaw AD, Shi J, Lopez MG, O’Neal JB, Billings FT IV. Intraoperative prediction of cardiac surgery-associated acute kidney injury using urinary biomarkers of cell cycle arrest. J Thorac Cardiovasc Surg. 2019;157(4):1545-1553.e5. doi: 10.1016/j.jtcvs.2018.08.090 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r26] 26.Kashani K, Al-Khafaji A, Ardiles T, et al. Discovery and validation of cell cycle arrest biomarkers in human acute kidney injury. Crit Care. 2013;17(1):R25. doi: 10.1186/cc12503 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r27] 27.Ely EW, Inouye SK, Bernard GR, et al. Delirium in mechanically ventilated patients: validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU). JAMA. 2001;286(21):2703-2710. doi: 10.1001/jama.286.21.2703 [DOI] [PubMed] [Google Scholar]

[soi240053r28] 28.Lindroth H, Khan BA, Carpenter JS, et al. Delirium severity trajectories and outcomes in ICU patients. defining a dynamic symptom phenotype. Ann Am Thorac Soc. 2020;17(9):1094-1103. doi: 10.1513/AnnalsATS.201910-764OC [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r29] 29.Katzman R, Brown T, Fuld P, Peck A, Schechter R, Schimmel H. Validation of a short Orientation-Memory-Concentration Test of cognitive impairment. Am J Psychiatry. 1983;140(6):734-739. doi: 10.1176/ajp.140.6.734 [DOI] [PubMed] [Google Scholar]

[soi240053r30] 30.Billings FT IV, Hendricks PA, Schildcrout JS, et al. High-dose perioperative atorvastatin and acute kidney injury following cardiac surgery: a randomized clinical trial. JAMA. 2016;315(9):877-888. doi: 10.1001/jama.2016.0548 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r31] 31.Bauer DF. Constructing confidence sets using rank statistics. J Am Stat Assoc. 1972;67(339):687-690. doi: 10.1080/01621459.1972.10481279 [DOI] [Google Scholar]

[soi240053r32] 32.Fagerland MW, Lydersen S, Laake P. Recommended confidence intervals for two independent binomial proportions. Stat Methods Med Res. 2015;24(2):224-254. doi: 10.1177/0962280211415469 [DOI] [PubMed] [Google Scholar]

[soi240053r33] 33.Althouse AD. Adjust for multiple comparisons? it’s not that simple. Ann Thorac Surg. 2016;101(5):1644-1645. doi: 10.1016/j.athoracsur.2015.11.024 [DOI] [PubMed] [Google Scholar]

[soi240053r34] 34.Feise RJ. Do multiple outcome measures require P value adjustment? BMC Med Res Methodol. 2002;2:8. doi: 10.1186/1471-2288-2-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r35] 35.Rosner MH, Okusa MD. Acute kidney injury associated with cardiac surgery. Clin J Am Soc Nephrol. 2006;1(1):19-32. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=17699187. doi: 10.2215/CJN.00240605 [DOI] [PubMed] [Google Scholar]

[soi240053r36] 36.McGuinness SP, Parke RL, Drummond K, et al. ; SO-COOL investigators . A multicenter, randomized, controlled phase IIB trial of avoidance of hyperoxemia during cardiopulmonary bypass. Anesthesiology. 2016;125(3):465-473. doi: 10.1097/ALN.0000000000001226 [DOI] [PubMed] [Google Scholar]

[soi240053r37] 37.Nam K, Nam JS, Kim HB, et al. ; Cardiac Surgery and Oxygen Therapy (CARROT) Investigators . Effects of intraoperative inspired oxygen fraction (FiO₂ 0.3 vs 0.8) on patients undergoing off-pump coronary artery bypass grafting: the CARROT multicenter, cluster-randomized trial. Crit Care. 2023;27(1):286. doi: 10.1186/s13054-023-04558-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r38] 38.Shaefi S, Shankar P, Mueller AL, et al. Intraoperative oxygen concentration and neurocognition after cardiac surgery. Anesthesiology. 2021;134(2):189-201. doi: 10.1097/ALN.0000000000003650 [DOI] [PMC free article] [PubMed] [Google Scholar]

[soi240053r39] 39.Holse C, Aasvang EK, Vester-Andersen M, et al. ; VIXIE Trial Group . Hyperoxia and antioxidants for myocardial injury in noncardiac surgery: a 2 × 2 factorial, blinded, randomized clinical trial. Anesthesiology. 2022;136(3):408-419. doi: 10.1097/ALN.0000000000004117 [DOI] [PubMed] [Google Scholar]

PERMALINK

Intraoperative Oxygen Treatment, Oxidative Stress, and Organ Injury Following Cardiac Surgery

Marcos G Lopez, MD, MS

Matthew S Shotwell, PhD

Cassandra Hennessy, MS

Mias Pretorius, MBChB

David R McIlroy, MBBS

Melissa J Kimlinger, MD

Eric H Mace, MD

Tarek Absi, MD

Ashish S Shah, MD

Nancy J Brown, MD

Frederic T Billings IV, MD, MSc

Key Points

Question

Findings

Meaning

Abstract

Importance

Objective

Design, Setting, and Participants

Interventions

Main Outcomes and Measures

Results

Conclusions

Trial Registration

Introduction

Methods

Trial Design and Oversight

Participants

Randomization and Masking

Treatment

Perioperative Participant Monitoring and Care

Outcomes and Safety Evaluation

Statistical Analyses

Results

Figure 1. CONSORT Diagram.

Table 1. Participant Characteristics.

Intraoperative Oxygenation

Table 2. Intraoperative Oxygenation.

Figure 2. Intraoperative Oxygenation Metrics in Participants Assigned to Hyperoxia and Normoxia.

Outcomes

Figure 3. Clinical Outcomes and Safety Events.

Safety End Points

One-Year Follow-up

Discussion

Limitations

Conclusions

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases