CABG indicates coronary artery bypass graft; CPB, cardiopulmonary bypass; ITT, intention to treat; and VF, ventricular fibrillation. The number of patients screened for eligibility was not available. “Stone heart” is a lethal complication that corresponds to postbypass ischemically-induced myocardial contracture.
aTreatment allocation was not transmitted as a result of a communication failure with the website.
CPB indicates cardiopulmonary bypass; ICU, intensive care unit. The size of the data markers reflects the precision of the estimation. The error bars indicate 95% CIs. Mechanical assist devices consisted of 22 intra-aortic balloon pumps in each group and 8 venoarterial extracorporeal life supports (5 in the levosimendan group and 3 in the placebo group). The decision for renal replacement therapy was at the discretion of the physicians in charge.
aTwo patients in the placebo group who refused participation after randomization could not be assessed for the primary end point and were excluded from the intention-to-treat analysis.
CABG indicates coronary artery bypass graft; LVEF, left ventricular ejection fraction. The size of the data markers reflects the precision of the estimation. The error bars indicate 95% CIs. P values for subgroup comparisons correspond to test for interaction.
Protocol for LICORN.
eFigure 1: Forest plot of relative risk differences for the primary end point and its components.
eFigure 2: Forest plot of the absolute risk difference in the primary end point according to predefined subgroups.
eFigure 3: Kaplan-Meier estimates of the probability of survival at Day 180.
eFigure 4: Cumulative incidence of patients “out of ICU and alive” at Day 28.
eFigure 5: Cumulative incidence of patients “off ventilator” at Day 28.
eTable 1 Euroscore II.
eTable 2: Relative Risks for in-hospital mortality, mortality at Day 28 and Day 180.
eTable 3: Complete list of serious adverse events in the levosimendan and placebo groups.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Cholley B, Caruba T, Grosjean S, Amour J, Ouattara A, Villacorta J, Miguet B, Guinet P, Lévy F, Squara P, Aït Hamou N, Carillon A, Boyer J, Boughenou M, Rosier S, Robin E, Radutoiu M, Durand M, Guidon C, Desebbe O, Charles-Nelson A, Menasché P, Rozec B, Girard C, Fellahi J, Pirracchio R, Chatellier G. Effect of Levosimendan on Low Cardiac Output Syndrome in Patients With Low Ejection Fraction Undergoing Coronary Artery Bypass Grafting With Cardiopulmonary BypassThe LICORN Randomized Clinical Trial. JAMA. 2017;318(6):548–556. doi:10.1001/jama.2017.9973
Does a prophylactic levosimendan infusion reduce the incidence of postoperative low cardiac output syndrome in patients with impaired left ventricular function who are undergoing isolated or combined coronary artery bypass grafting surgery under cardiopulmonary bypass?
In this randomized clinical trial involving 335 patients, levosimendan compared with placebo did not result in a significant difference in the composite end point of prolonged catecholamine infusion, use of left ventricular mechanical assist device, or renal replacement therapy (52% in the levosimendan group vs 61% in the placebo group).
Levosimendan was not effective in reducing the incidence of postoperative low cardiac output syndrome in patients such as these.
Low cardiac output syndrome after cardiac surgery is associated with high morbidity and mortality in patients with impaired left ventricular function.
To assess the ability of preoperative levosimendan to prevent postoperative low cardiac output syndrome.
Design, Setting, and Participants
Randomized, double-blind, placebo-controlled trial conducted in 13 French cardiac surgical centers. Patients with a left ventricular ejection fraction less than or equal to 40% and scheduled for isolated or combined coronary artery bypass grafting with cardiopulmonary bypass were enrolled from June 2013 until May 2015 and followed during 6 months (last follow-up, November 30, 2015).
Patients were assigned to a 24-hour infusion of levosimendan 0.1 µg/kg/min (n = 167) or placebo (n = 168) initiated after anesthetic induction.
Main Outcomes and Measures
Composite end point reflecting low cardiac output syndrome with need for a catecholamine infusion 48 hours after study drug initiation, need for a left ventricular mechanical assist device or failure to wean from it at 96 hours after study drug initiation when the device was inserted preoperatively, or need for renal replacement therapy at any time postoperatively. It was hypothesized that levosimendan would reduce the incidence of this composite end point by 15% in comparison with placebo.
Among 336 randomized patients (mean age, 68 years; 16% women), 333 completed the trial. The primary end point occurred in 87 patients (52%) in the levosimendan group and 101 patients (61%) in the placebo group (absolute risk difference taking into account center effect, −7% [95% CI, −17% to 3%]; P = .15). Predefined subgroup analyses found no interaction with ejection fraction less than 30%, type of surgery, and preoperative use of β-blockers, intra-aortic balloon pump, or catecholamines. The prevalence of hypotension (57% vs 48%), atrial fibrillation (50% vs 40%), and other adverse events did not significantly differ between levosimendan and placebo.
Conclusions and Relevance
Among patients with low ejection fraction who were undergoing coronary artery bypass grafting with cardiopulmonary bypass, levosimendan compared with placebo did not result in a significant difference in the composite end point of prolonged catecholamine infusion, use of left ventricular mechanical assist device, or renal replacement therapy. These findings do not support the use of levosimendan for this indication.
EudraCT Number: 2012-000232-25; clinicaltrials.gov Identifier: NCT02184819
Mortality after coronary artery bypass grafting (CABG) surgery declined from 2.4% to less than 1.5% between 1994 and 2009, but remained much higher (17% to 24%) in the subgroup of patients who developed a postoperative low cardiac output syndrome.1 The prevalence of low cardiac output syndrome after CABG surgery varies from 3% to 14%, but the presence of a preoperative left ventricular dysfunction, ie, a left ventricular ejection fraction less than or equal to 40%, is associated with an odds ratio of 2.0 (95% CI 1.7-2.4).1Quiz Ref ID The low cardiac output syndrome is also associated with more frequent complications, including pulmonary impairment, myocardial infarction, stroke, and renal failure.2
The therapeutic management of low cardiac output syndrome involves inotropic agents as first-line treatment and mechanical assist devices when pharmacologic support fails to restore adequate perfusion. However, inotropic agents have been associated with an increase in morbidity and mortality after cardiac surgery.3,4 Therefore, new drugs with less deleterious adverse effects could be useful. Several studies and meta-analyses have suggested that levosimendan (Orion Pharma) could help prevent low cardiac output syndrome and reduce morbidity and even mortality after cardiac surgery.5-13 However, 2 recent, large, randomized, clinical trials did not confirm these findings when levosimendan was used prophylactically, ie, before cardiac surgery,14 or implemented in the presence of an already manifest postoperative low cardiac output syndrome.15 The Levosimendan in Coronary Artery Revascularization trial was conducted during the same period as these 2 studies.
The aim of this trial was to assess whether a 24-hour levosimendan infusion started at anesthetic induction in patients with impaired left ventricular function (left ventricular ejection fraction ≤40%) who were undergoing CABG (alone or combined) with the use of cardiopulmonary bypass could reduce the prevalence and severity of postoperative low cardiac output syndrome.
The trial was approved by an independent National Research Ethics Committee (CPP Ile de France VI). The study protocol was published previously16 and is available in Supplement 1. All patients provided written informed consent to participate.
Thirteen cardiac surgical centers participated in this study. Patients scheduled for CABG surgery with cardiopulmonary bypass alone or combined with valve surgery procedure were eligible if they had a left ventricular ejection fraction of 40% or lower. Patients undergoing isolated valve surgery were not eligible. Exclusion criteria were younger than 18 years, pregnancy, renal failure (creatinine clearance <30 mL/min), liver failure (prothrombin ratio <50%), emergency surgery, hypotension (mean arterial pressure <60 mm Hg), or tachycardia (heart rate >120/min). Patients could be enrolled if they were receiving preoperative inotropes or if the surgeon decided to insert an intra-aortic balloon pump as a prophylactic measure before cardiopulmonary bypass.
This trial was a randomized, double-blind, 2-arm, parallel-group, placebo-controlled study (Figure 1). Quiz Ref IDPatients were assigned to receive prophylactic levosimendan or placebo in a 1:1 ratio with a deterministic dynamic allocation.17 The procedure took into account the center, the surgery performed (CABG alone or combined surgery), left ventricular ejection fraction (<30%, or between 30% and 40%), preoperative treatment by inotropes or intra-aortic balloon pump, and preoperative β-blocker therapy. This minimization procedure was expected to balance the distribution of prognostic factors between the 2 groups and within each center. A random component was introduced to reduce the predictability of allocation. An independent statistician established the randomization algorithm, which was installed on a website with secure access (as described in the protocol (Supplement 1).16
We stored the study drugs at 5°C±3°C until reconstitution with 500 mL of 5% dextrose solution immediately before administration to the patient. The placebo was made of riboflavin to reproduce the yellow color of levosimendan and was indistinguishable from the intervention drug. No loading dose was administered. A continuous prophylactic infusion of 0.1 µg/kg/min was started just after anesthetic induction and was maintained over 24 hours in the absence of serious adverse reactions. Study drug administration was discontinued in the event of an anaphylactic reaction, refractory hypotension (defined as a mean arterial pressure <60 mm Hg despite optimal therapeutic management at the discretion of the anesthesiologist), or intractable arrhythmias.
Quiz Ref IDThe primary end point was a composite of 3 elements reflecting low cardiac output syndrome: catecholamine (ie, dobutamine, epinephrine, norepinephrine, or milrinone) infusion persisting beyond 48 hours after the initiation of the study drug, the need for circulatory mechanical assist devices in the postoperative period (for patients who had an intra-aortic balloon pump inserted preventively, it was considered that they met the outcome measure if they were not weaned from the balloon within 96 hours after study treatment initiation), or the need for renal replacement therapy at any time during intensive care unit stay. If a patient had at least 1 of these criteria, he or she was considered as meeting the primary end point.
The secondary end points were in-hospital mortality, mortality at days 28 and 180, each component of the primary end point, number of days with circulatory mechanical assist device, number of days with catecholamine infusion, number of days with renal replacement therapy and number of renal replacement therapy kits (ie, disposable circuits and filters) that were used for each patient, and number of ventilator-free days, out-of-intensive care unit and out-of-hospital days at day 28, and total hospital and intensive care unit lengths of stay. The study drug safety was assessed by recording of the incidence of hypotension, defined as mean arterial pressure less than 60 mm Hg occurring after study drug initiation. Data on fluids (volume and type of fluid) or vasopressors used to reverse hypotension, as well as arrhythmias, conduction disturbances, and the treatments they required, were also recorded. Refractory hypotension was defined as a failure to increase mean arterial pressure above 60 mm Hg despite optimal management. The myocardial damage was reflected by the highest troponin value measured between hours 24 and 48.
All adverse events and serious adverse events were prospectively recorded and reported to an independent data and safety monitoring board, which had full access to unblinded data and periodically reviewed the safety results.
The number of patients required was estimated to be 340 (170 per group) according to Nquery Advisor version 4.0 statistical software, with the following assumptions: (1) prevalence of the primary end point in the control group: 65% (proportion observed in the same population between 2002 and 2004 at center 1; ); (2) prevalence of the primary end point in the levosimendan group: 50% (ie, 15% absolute risk difference, more conservative than that observed by Levin et al18 but attesting for a clinical benefit); and α level of 5% and power of 80%.
The primary analysis was performed on an intention-to-treat basis for all patients included in the trial except 3 (1 withdrew consent and 2 refused surgery, declined further participation, and had no criteria of judgment available). We analyzed the composite primary end point and its components according to the following strategy: first, absolute risk differences were estimated by fitting a binomial regression model using the identity link function, including a random effect to account for center effect. In case of nonconvergence of the model, a binomial regression model with a robust variance was used. Second, relative risks were estimated by fitting a log-binomial regression model. In case of nonconvergence of the model, a log-binomial regression model with a robust variance was used. Mortality data were analyzed with the same methods used for the primary end point. The other secondary end points were analyzed as count data with a Poisson model including a random effect to take into account center effect. In case of overdispersion, a negative binomial distribution was used. The log-rank test was used to compare survival curves. Cumulative incidences were compared with the Gray test.
Planned subgroup analyses assessed whether the effect of levosimendan varied according to type of surgical procedure performed (CABG alone or combined surgery); left ventricular ejection fraction less than 30%, or 30% to 40%; preoperative catecholamine or mechanical assist device; and preoperative β-blockers. The heterogeneity of the treatment effect was evaluated by an interaction test and is presented with a forest plot.19 The P value for interaction was obtained from the log-binomial model.
For all hypotheses testing, 2-sided tests were used and P < .05 was considered to indicate statistical significance. No adjustment was made for multiple comparisons; therefore, all secondary outcomes should be considered exploratory. Because of the small amount of missing data, multiple imputation was not performed. All calculations were performed with SAS version 11.3.
A total of 336 patients were eligible and randomized from June 2013 until May 2015 and followed up during 6 months (last follow-up November 30, 2015). A total of 333 patients completed the trial, as detailed in Figure 1. Vital status was missing for 6 patients. Levosimendan and placebo groups were well balanced: mean (SD) age (69  vs 67  years), male sex (83% vs 85%), mean (SD) ejection fraction (33% [6%] vs 33% [6%]), and Euroscore II (3.1% vs 3.4%), respectively. The baseline characteristics are presented in Table 1 and the surgical procedures are described in Table 2. In both groups, 74% of patients had isolated CABG and 26% had combined surgery. The proportion of patients with left ventricular ejection fraction of 30% or lower was 20% in the levosimendan group and 23% in the placebo group (P = .44). The complete details of the Euroscore II are presented in eTable 1 in Supplement 2). Ninety-one percent of the participants received an infusion of study drug during 24 hours (SD, 3 hours).
In the intention-to-treat population, the primary end point occurred in 87 levosimendan patients (52%) and 101 placebo patients (61%) (absolute risk difference, −7%; 95% CI, −17% to 3%; P = .15) (Figure 2). Among the 3 components of the primary end point, the need for catecholamine infusion beyond 48 hours of study drug initiation was the most frequent (n = 180; 54%) but was not significantly affected by levosimendan administration, with an absolute risk difference of −8% (95% CI, −18% to 1%; P = .09). The need for mechanical assist device (or failure to wean from intra-aortic balloon pump at hour 96 of study drug initiation, when the balloon was inserted preoperatively) and the need for renal replacement therapy were also not affected by levosimendan (Figure 2). In the per-protocol population, the primary end point occurred in 81 levosimendan patients (51%) and 95 placebo patients (61%) (absolute risk difference, −8%; 95% CI, −18% to 3%; P = .14). Predefined subgroup analysis found no interaction with left ventricular ejection fraction lower than 30%, preoperative treatment with β-blockers, type of surgery, or preoperative intra-aortic balloon pump or preoperative catecholamine infusion (Figure 3). Relative risks in regard to the primary end point and its components, as well as for the predefined subgroups, are presented in the supplement (eFigures 1 and 2 in Supplement 2, respectively).
No significant difference was found in any secondary end point between the 2 groups, including mortality at days 28 and 180 (Table 3). Absolute risk difference for mortality at day 180 was −1% (95% CI, −6% to 3%; P = .59). Relative risks for mortality and the actuarial survival curve are presented in eTable 2 and eFigure 3 in Supplement 2, respectively. Cumulative incidence curves representing probability of intensive care unit discharge at day 28 are shown in eFigure 4 in Supplement 2 and weaning from mechanical ventilation at day 28 in eFigure 5 in Supplement 2.
There was no significant difference in the incidence of adverse events or serious adverse events between the 2 groups (Table 4). Severe hypotension occurred in 57% of patients receiving levosimendan and 48% of those receiving placebo (P = .11). The management of hypotension was comparable between the 2 groups, except for the use of epinephrine. Refractory hypotension requiring study drug interruption occurred in 5 patients from the levosimendan group vs 1 from the placebo group (P = .12). There was no significant between-group difference in the incidence of postoperative atrial fibrillation (50% vs 40%, levosimendan vs placebo, respectively; P = .09), or third-degree atrioventricular blocks (4% vs 9%, respectively; P = .08). During the hospital stay, the use of antiarrhythmic agents was comparable between the 2 groups, but more levosimendan patients received amiodarone in the operating room (19 [11%] vs 6 [4%] placebo patients; P = .007). The peak troponin value measured between hours 24 and 48 was not significantly different between the 2 groups (Table 4).
Quiz Ref IDIn this randomized trial of patients with ejection fraction of 40% or lower who were undergoing CABG with cardiopulmonary bypass, levosimendan compared with placebo did not result in a significant difference in the composite end point of prolonged catecholamine infusion, use of left ventricular mechanical assist device, or renal replacement therapy.
The low cardiac output syndrome is a postoperative complication defined as the need for catecholamine infusion or mechanical assist device to maintain adequate arterial blood pressure and cardiac index.20 Patients with low cardiac output syndrome also have a high rate of renal failure (≈ 20%) as a consequence of inadequate tissue perfusion.21,22 Therefore, if a patient does not require prolonged postoperative catecholamine infusion, mechanical assist device, or renal replacement therapy, it is unlikely that he or she has experienced low cardiac output syndrome. This is why the combination of these 3 clinical markers was chosen as a simple and robust primary end point to evaluate the efficacy of levosimendan.
Administering inotropic support to patients with impaired left ventricular ejection fraction at cardiopulmonary bypass weaning is common even if there is no measurement attesting for a low cardiac output. However, these infusions are usually tapered and stopped quickly in the intensive care unit when there is no persistent cardiac dysfunction, whereas their maintenance beyond 48 hours postoperatively reflects a true impairment in myocardial contractility. Inotropic support at hour 48 was required in 49% of levosimendan patients as opposed to 59% of those who received placebo (P = .09).
These findings are consistent with those of Mehta et al14 (the LEVO-CTS study [Levosimendan in Patients with Left Ventricular Systolic Dysfunction Undergoing Cardiac Surgery Requiring Cardiopulmonary Bypass]), who recently reported, in a similar cohort of patients in whom levosimendan infusion was started at anesthetic induction, a lack of significant reduction in their primary end point, composed of mortality at day 30; the need for mechanical assist device at day 5; the need for renal replacement therapy through day 30; and myocardial infarction through day 5. Likewise, in another recent randomized trial (the CHEETAH study [Levosimendan to Reduce Mortality in High Risk Cardiac Surgery Patients: A Multicenter Randomized Controlled Trial]), Landoni et al15 did not find a significant difference in mortality at day 30 (primary end point) in their population, in which levosimendan was administered only after the onset of severe cardiac dysfunction after cardiopulmonary bypass.
None of the predefined factors (prophylactic intra-aortic balloon pump or catecholamine support, preoperative β-blockers, severity of left ventricular ejection fraction alteration, or type of surgery) was found to influence the effect of levosimendan. The lack of significant difference according to preoperative β-blockade does not corroborate previous observations.23 To explain this discrepancy, it is likely that β-blockers were not reintroduced in the early postoperative period in patients with hemodynamic compromise who were treated with the β-agonist dobutamine.
No significant difference in mortality (up to day 180) was found between the levosimendan and placebo-treated groups, a result in sharp contradiction with the conclusions of several meta-analyses11-13 but confirming the findings of the 2 abovementioned recent large randomized clinical trials.14,15 Similarly, the need for renal replacement therapy was not reduced by the use of levosimendan in this study, as well as in the LEVO-CTS and CHEETAH trials, a finding that contradicts results of previous studies supporting the improvement in renal function with levosimendan after cardiac surgery.9,10,12,24,25
The total numbers of adverse events and serious adverse events were not significantly different between the 2 groups. Refractory hypotension leading to study drug interruption was observed in 5 patients in the levosimendan group and 1 in the placebo group (P = .12). Nonsignificant excess in hypotension was also observed in the LEVO-CTS trial, in which patients received the same regimen of levosimendan as that used in this study, but not in the CHEETAH trial, in which the average dose was 0.07 µg/kg/min. The prevalence of postoperative atrial fibrillation was not significantly higher in the levosimendan group (50% vs 40%; P = .09), but an excess of postoperative atrial fibrillation has been observed in medical patients with advanced left ventricular heart failure,26 as well as in surgical patients after cardiac procedures.27 The role of levosimendan in protecting the myocardium was not specifically addressed in this study, but no difference was observed in peak troponin values, unlike what had been reported previously.7,28 This lack of clinically relevant anti-ischemic effect was also suggested by the 2 other randomized clinical trials that found no difference in postoperative myocardial infarction.14,15
Quiz Ref IDThis study has several limitations. First, the use of a small dose of levosimendan without an initial bolus was deliberately chosen to avoid an excess of vasodilatation in this fragile population, as currently recommended.29 This might explain, in part, the minimal effect of levosimendan observed in this study. Second, the timing of administration was also very close to the “myocardial injury” resulting from cardiopulmonary bypass and cardioplegic arrest, thus limiting the potential protective effect of the drug related to the opening of cardiac mitochondrial adenosine triphosphate–dependent potassium channels.30 Starting levosimendan several hours before surgery would have required an environment that was not available routinely at most institutions to monitor the effect of the infusion. Third, the end point of the study involved decisions that may vary according to the center and to caregivers, which is why we performed statistical comparisons with models taking into account the variability related to the center. Fourth, this study failed to enroll the anticipated number of participants by 5 (335 instead of 340) because recruitment had to be stopped when study drugs reached their expiration date. However, this small gap is unlikely to have severely reduced the anticipated power of the study, especially because only 2 patients had no data on the primary end point and had to be removed from the intention-to-treat analysis. The number of patients to include in this trial was estimated according to an expected prevalence of the composite primary end point of 65% in the target population. The observed prevalence was 61% in the placebo group, attesting that the initial assumption was realistic. We hypothesized that levosimendan would reduce this figure to 50%, but the observed prevalence was 52% in the intention-to-treat population (51% in the per-protocol population). Because the study was powered according to a 15% absolute risk reduction and the point estimate favored levosimendan (by 7%) but included a reduction of 17%, it is possible that the study was underpowered to definitely rule out a clinically important benefit for the drug for the primary composite outcome.
Among patients with low ejection fraction who were undergoing CABG with cardiopulmonary bypass, levosimendan compared with placebo did not result in a significant difference in the composite end point of prolonged catecholamine infusion, use of left ventricular mechanical assist device, or renal replacement therapy. These findings do not support the use of levosimendan for this indication.
Corresponding Author: Bernard Cholley, MD, PhD, Department of Anesthesiology and Critical Care Medicine, Hôpital Européen Georges Pompidou, AP-HP, 20 rue Leblanc, 75908 Paris Cedex 15, France (email@example.com).
Accepted for Publication: July 6, 2017.
Author Contributions: Drs Cholley and Caruba had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Cholley, Caruba, Chatellier.
Acquisition, analysis, or interpretation of data: Cholley, Caruba, Grosjean, Villacorta, Lévy, Ait Hamou, Carillion, Boughenou, Rosier, Durand, Guidon, Desebbe, Charles-Nelson, Menasché, Rozec, Fellahi, Pirracchio, Chatellier.
Drafting of the manuscript: Cholley, Caruba, Charles-Nelson, Pirracchio, Chatellier.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Charles-Nelson, Pirracchio, Chatellier.
Obtained funding: Cholley, Caruba.
Administrative, technical, or material support: Caruba, Villacorta, Lévy, Ait Hamou, Boughenou, Rosier, Guidon, Desebbe, Rozec, Fellahi, Chatellier.
Supervision: Cholley, Caruba.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Cholley reports receiving an honorarium for an invited lecture for Orion Pharma. Dr Menasché reports serving on the advisory boards of Edwards Lifesciences and Gecko Biomedical. Dr Rozec reports receiving a grant from Baxter and receiving personal fees from Baxter and Xenios. No other disclosures were reported.
Funding/Support: The LICORN study was funded by the French Ministry of Health (Programme Hospitalier de Recherche Clinique national 2011, MIN02-07) and sponsored by Assistance Publique-Hôpitaux de Paris. Orion Pharma provided study drugs free of charge.
Role of the Funder/Sponsor: Orion Pharma and the other 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.
LICORN Collaborators: Hôpital Européen Georges Pompidou, AP-HP, and Université Paris Descartes-Sorbonne Paris Cité, Paris, France: Christian Latrémouille, MD, PhD; Paul Achouh, MD, PhD; Jérôme Jouan, MD; Alain Bel, MD; Jean-Noël Fabiani, MD, PhD; Delphine Hourton, PhD; Centre Hospitalo-Universitaire de Dijon-Bourgogne: Olivier Bouchot, MD, PhD; Hôpital de La Pitié Salpêtrière, AP-HP, and Université Pierre et Marie Curie: Adrien Bouglé, MD; Astrid Quessard, MD; Marwan Nader, MD; Pascal Leprince, MD, PhD; CHU de Bordeaux: Delphine Hirtz, MD; Alain Coiffic, MD; Nathalie Noël; Cécile Poisvert; Alain Rémy, MD; Cédric Zaouter, MD; Louis Labrousse MD, PhD; Laurent Barandon, MD, PhD; CHU La Timone, Marseille: Françoise Gaillat, MD; Catherine-Charlotte Joseph, MD; Frédéric Collart, MD, PhD; Hôpital Laënnec, Nantes: Nicolas Cotron, MD; Laurent Delille, MD; Jean-Christian Roussel, MD, PhD; Hubert-François Carton, MD; Hôpital Pontchaillou, Rennes: Laurent Daviet, MD; Erwan Flecher, MD, PhD; Nouvel Hôpital Civil, Strasbourg: Minh Tam Hoang, MD; Clinique Ambroise Paré, Neuilly: Alain Brusset, MD; Philippe Estagnaisie, MD; Hôpital Bichat, AP-HP, Paris: Dan Longrois, MD, PhD; Hôpital Lariboisière, AP-HP, Paris: Alexandre Mebazaa, MD, PhD; Hôpital Claude Huriez, Lille: Olivier Joulin, MD; Grenoble: Sylvaine Robin, MD, Géraldine Dessertaine, MD; Myriam Cassez Brasseur, MD; Olivier Chavanon, MD, PhD; CHU Côte de Nacre, Caen: Fabien Dechanet, MD; Clément Boisselier, MD; Hôpital Cardiologique Louis Pradel, Lyon: Pierre Joseph, MD; Olivier Bastien, MD, PhD; Jean-François Obadia, MD, PhD.
Data and Safety Monitoring Board: Hôpital Henri Mondor, AP-HP, Créteil: Pascal Gueret, MD, PhD; Hôpital Marie Lannelongue, Le Plessis-Robinsson: François Stéphan, MD, PhD; Hôtel-Dieu, AP-HP, Paris, INSERM 1153: Raphaël Porcher, MD, PhD.
Meeting Presentations: Presented at: Meeting of the French Society of Anesthesiologists, September 24, 2016, Paris, France; the International Symposium on Intensive Care and Emergency Medicine, March 23, 2017, Brussels, Belgium; and the European Association of Cardiothoracic Anaesthesiology meeting; April 20, 2017; Berlin, Germany.
Additional Contributions: We thank the team of the Clinical Research Department of Hôpital Européen Georges Pompidou, AP-HP, for expert study management and the Association des Anesthésistes-Réanimateurs en Chirurgie Cardio-Thoracique & Vasculaire (ARCOTHOVA) for facilitating the organization of this study by connecting several French cardiac surgical centers.