Perioperative Individualized Optimization of Mean Arterial Pressure in Cardiac Surgery

Neurologic and renal complications frequently occur after cardiac surgery. Acute renal failure following cardiac surgery increase the risk of chronic kidney disease, while postoperative neurological complications increased the risk of chronic cognitive dysfunction. Many cardiac surgical patients suffer from systemic hypertension, but the goal in clinical practice is to maintain the mean arterial pressure (MAP) above 65 mmHg. The investigators test the hypothesis that an individualized MAP optimization during the per-operative and the 24 hours postoperative period should decrease the renal and neurological complications following cardiac surgery.

The investigators propose a randomized controlled study conducted in 21 French cardiac surgical centers.

Inclusion criteria :

• Age > 18 yr
• Patient scheduled for cardiac surgery with cardiopulmonary bypass (aortic valve repair or replacement and/or aortic surgery with normothermia and/or coronary artery bypass)
• Patient able to understand and voluntarily sign an informed consent form
• Patient able to adhere to the study visit schedule and other protocol requirements
• Patient affiliated with an appropriate social security system.
• French speaking patient

Exclusion criteria :

• Glomerular filtration rate \< 30 ml/min/1.73m2
• Neurologic disorder (motor and/or sensory deficit, cognitive disorder)
• Another type of surgery (emergency surgery, mitral or tricuspid repair or replacement, congenital surgery)
• Left ventricular ejection fraction \< 30% or acute heart failure in the month before surgery
• Pulmonary artery pressure > 60 mmHg
• Endocarditis
• Surgery with hypothermia (\< 34°C)
• Hepatic cirrhosis
• Alcohol use disorder (Score AUDIT C ≥ 5)
• Refusal to consent or adults with protective measures (curatorship or tutorship or safeguarding justice or juridical protection)
• Pregnancy, or breast-feeding

Protocol :

Included patients will be allocated during the peroperative and 24hours postoperative periods to:

• individualized mean arterial pressure group

• mean arterial pressure above 65 mmHg In both groups,

  • The volemia will be optimized (least to the discretion to the attending anesthesiologist and intensivist) before vasopressor prescribing with ephedrine at first (until 30mg), the norepinephrine will be used then after if the MAP is not in the range.
  • The output of cardiopulmonary bypass will be adapted using following targets:
  • SV02 > 70%,
  • DO2 > 280 ml/min/m2 (using a dedicated D02 monitoring),
  • Hb > 7.5 g/dL.

Primary judgment criteria

• Incidence of patients (percentage and 95% confidence interval) in each treatment group suffered of at least mortality, acute renal insufficiency (KDIGO classification ≥ 1) and/or neurological complication (stroke, confusion,), during the 7 days following cardiac surgery.

Secondary judgment criteria

• Incidence of mortality (percentage and 95% confidence interval) in each treatment group during the 7 days following cardiac surgery and 90 days +/-10 days after surgery.
• Incidence of patients (percentage and 95% confidence interval) suffered from neurological complications (confusion using CAM-ICU scale and/or stroke defined by neurological clinical deficit with a cerebral tomodensitometry confirmation) in each treatment group during the 7 days following cardiac surgery and 90 days +/- 10 days after surgery. Trained care providers or trained research staff member will assess patients for confusion (primary outcome) using the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU). This method (the CAM-ICU) have been shown to be reliable and to have good agreement with the DSM-V criteria (see appendix 1).

Measurements will start from postoperative day 0 after the end of surgery and the end of sedative drugs to day 7 once daily.

The CAM-ICU algorithm consists of 4 items (see appendix 1): 1. Acute Onset or Fluctuating Course. 2. Inattention 3. Disorganized thinking. 4. Altered Level of consciousness. The diagnosis of confusion by CAM-ICU requires a positive response to features 1 and 2 plus either 3 or 4.

• Incidence of patients (percentage and 95% confidence interval) suffered from acute renal insufficiency (KDIGO classification ≥ 1) in each treatment group during the 7 days following cardiac surgery and 90 days +/- 10 days after surgery (Appendix 2).
• Quality of life using EQ5D-3L test at POD7 and POD90 days +/- 10 after surgery, in comparison with the preoperative evaluation (see appendix 3).
• Post-operative cognitive dysfunction. The MoCA test (see appendix 4) will be used to screen patients' pre-operative cognitive functions at the preanaesthetic consultation. This test explores a representative panel of cognitive functions in 10 minutes and exhibits higher sensitivity and specificity than the Mini-Mental State. Cognitive functions are considered normal if MoCA >26. MoCA between 20 and 26 is considered mild cognitive dysfunction, and MoCA\<20 is considered severe cognitive dysfunction. The tests for detecting POCD should be based on well-described sensitivity and suitability in relation to surgical patients. The evaluation should be based on differences between pre- and postoperative performance (POD7 and 90 days +/- 10 days after surgery).

Number of participants : 1100

Statistical analysis :

On the basis of data from trials that examined this population of patients undergoing cardiac surgery, the investigators hypothesized that 25% of the control group would have a mortality and/or neurologic and/or renal failure (primary endpoint) within 7-day of surgery (Appendix 1) With a two-sided alpha risk of 5% (corresponding to a one-sided 2.5% alpha risk), a power of 80% and a decrease of 7% of the primary endpoint in the experimental group (i.e. 18%), 542 patients per group are required (1082 in total, PROC SEQDESIGN in SAS 9.4). Given the short follow-up and the population studied, the investigators expect a very high proportion of patients who will complete the study. The investigators therefore plan to include 1100 patients in total. Because the effect size is largely unknown, the investigators created a two-sided two-stage group sequential design with early stopping to reject the null hypothesis (PROC SEQDESIGN in SAS 9.4). A single interim analyse will be performed after the inclusion of 550 patients (O'Brien\&Fleming plan).

Accordingly, the study will stop for efficacy if nominal z-value (corresponding P-values) is above 2.79651 (P\<0.0052) at the interim analysis (542 patients). This analysis will be conducted independently from investigators and presented to the Data Safety Monitoring Board. If the study goes to the final analysis (i.e. stage 2, 1082 patients), the P-value for statistical significance will be 0.048 corresponding to a nominal z-value of 1.97743.

A p-value below the error spending bounds will be considered as statistically significant (see above).

The statistical analysis will be done using SAS software version 9.4 (NC, Cary) by Prof. Jean-Jacques Parienti, at the Unit of Biostatistics and Clinical Research of the Caen University hospital, France. External statistical analysis could be performed if necessary.

A flow chart will describe screened, randomized and analyzed patients according to the CONSORT Figure. Baseline characteristics will be described by numbers (percentages), mean (Standard deviation) and median (interquartile range), as appropriate, according to their randomized group.

Regarding the analysis of the primary endpoint, the percentages of neurologic or renal failure between randomized groups will be compared with the use of a Cochran-Mantel-Haenszel chi-square test. The effect size of the experimental group will be quantified by the Cochran-Mantel-Haenszel Estimate for a Risk Ratio and its 95% confidence interval stratified on age group (\< versus ≥ 75 years old and combined versus single type of cardiac surgery), as appropriate.

Regarding the analysis of secondary endpoints, the first analysis will be to examine each component of the composite primary endpoint separately. This analysis will follow the same statistical plan as describe above. The comparison between groups for qualitative secondary endpoints will follow the same statistical plan than the primary endpoint and the comparison between groups for quantitative secondary endpoints will be computed by Student t-test or Mann-Whitney U test, as appropriate. The other secondary endpoints will be tested as exploratory.

Subgroup analysis based on the primary judgment criteria will be done. Factors used for the stratification of the randomization will be considered for these analyses, with Breslow-day test for interaction.

The full-set analysis will include all randomized subjects. The analyses will be carried out with intention to treat in the goal of minimising potential bias. In case of missing data for the primary outcome, multiple imputations will be performed to comply with the intent to treat approach and the complete case analysis will be performed as a sensitivity analysis.

A p-value below the O'Brien\&Flerming bounds will be considered as statistically significant (see above).

The statistical analysis will be done using SAS software version 9.4 (NC, Cary) by Prof. Jean-Jacques Parienti, at the Unit of Biostatistic and Clinical Research of the Caen University hospital, France; or another independent and external biostatistical expert.

Methods of managing digital data:

The data will be managed in a database administered by the sponsor. The data entry will be carried out by the investigators of each center and may be carried out by any person registered on the delegation of tasks list. The system used for the computerized database will be following the regulations in force, in particular the MR-01. Access to the database will be limited to authorized persons only (team promoter, investigation team, principal investigator) and the rights (reading, writing, specific pages of the CRF, patients of the center or all the patients, ...) will be attributed according to their role in the study and their center. The authentication of the users will be done using a username and a personalized password for each user. Connections to the database will be saved in the connection history.

Data validation checks may be scheduled according to the level of risk and/or the impact of the study defined by the sponsor. Following the execution of these tests, requests for clarification (queries) may be sent to the investigators to correct or confirm data entered in the database. Regarding the risk and/or the impact of the study, a pre-analysis committee may meet depending on the level of risk and/or the impact of the study defined by the promoter before exporting the data, to qualify the deviations identified during the study and decide on the inclusion or not of the data collected in the study.