Cardiac surgery is sometimes associated with organ dysfunction of variable severity (renal insufficiency, cognitive decline, arrhythmias, ARDS). The phenomenon responsible is an intense inflammatory reaction induced by cardiopulmonary bypass, leading to microcirculation alterations, specially in endothelial cell and its protective layer - glycocalyx. Endothelial dysfunction then reduces the reactivity of peripheral tissues to hypoxia, and is associated with bad prognosis.

High - dose corticoids administration at anesthesia induction in cardiac surgery could attenuate the intensity of this inflammatory reaction, and represents the current practice in our hospital. Nevertheless, this attitude is abandoned in numerous cardiac surgery centres.

Introduction Cardiac surgery is associated with variable severity organ dysfunction, which further leads to an increase of ICU length of stay (ICU LOS) or even higher mortality. Various mecanisms enter into account, from global through regional hemodynamics until microcirculation alterations.

These are due to an intense inflammatory reaction triggered by the contact of blood with cardiopulmonary bypass surfaces, the contact blood - air from cardiotomy aspiration, heparin and heparin - protamin complexes, ischemia - reperfusion mediators, and toxins from the digestive tube.

These alterations are observed in both the beating heart surgery and surgery with cardiopulmonary bypass (CBP). The inflammatory response is probably less important in off - pump surgery though this is hypothesis often questionned and some publications show that the changes associated with CBP could only be transient.

Inflammatory response is possibly reduced by high - dose corticoid administration at anesthesia induction - by preserving cell membrane integrity, leucocyte adhesion and cellular immunity attenuation, and finally by complement and cytokine release reduction.

Numerous metaanalyses show benefits of this strategy with an acceptable security profile:

• Atrial fibrillation reduction
• Possible ICU LOS reduction
• Same incidence of wound infection

Within the risks can be observed:

• Higher incidence of hyperglycemia
• No impact on overall mortality However, two recent big RCTs did not find any benefit as in terms of morbidity-mortality, incidence of wound infection and possibly an increase in number of reinterventions for tamponnade.

Providing the low level of evidence on real benefits and some of the secondary effects, the systematic use of corticoids (as applied in our centre) was abandonned by a some of cardiac surgery centres.

Nevertheless, corticoids could be beneficial in specific subgroups. In order to identify these patients, it is important to understand the mecanism of microcirculation alterations in the setting of cardiac surgery with CBP and the effect of corticoids on CBP induced inflammatory processes.

The alterations of endothelial function under CBP may be observed at three levels :

1. endothelial reactivity (vasodilation et vasoconstriction)
2. blood flow (acceleration et deceleration)
3. glycocalyx (endothelial protective layer) These three components of microcirculation are tightly linked. De Backer study on sublingual microcirculation shows its alterations specific for this type of surgery (more pronounced in cardiac surgery with CBP than off-pump) and not found in minor surgery (thyroidectomy). He also found a positive correlation with lactate level. Cardiac surgery can be considered as an interesting microcirculation and endothelial function study model.

The hypothesis of investigators is that microcirculation modifications along with those of endothelial function in cardiac surgery with CBP are due to endothelial cell impairement and impairement of its protective layer - glycocalyx, condition which deregulates the peripheral tissue reactivity to hypoxia, and that corticoids could reduce these alterations.

The investigators will study the endothelial function by three methods :

A/ FMD (Flow Mediated vasoDilation) which is a non invasive technique considered as the gold standard for endothelial reactivity evaluation since the end of 1990s. It is based on echographic measurement of brachial artery diameter variation and the Doppler flow evaluation, during an episode of transient ischemia induced by pneumatic cuff inflation (vascular occlusion test).

Three studies showed a reversible alteration of endothelial function mainly after cardiac surgery using CBP with continuous blood flow (FMD variation of about 11%). Expecting a reduction of endothelial dysfunction with corticoids by 5%, the investigators will analyze a group of 60 patients (30 in the corticoids group and 30 in the placebo group). In previous studies, variation is already significative when FMD varies from 1,5 - 2%.

B/ NIRS (near infrared spectroscopy) - it's a non invasive technique which measures peripheral tissue oxygen saturation and allows to assess oxygen consumption and microcirculation reactivity during an arterial occlusion test distally to compressive cuff.

Alterations of the variables derived from delta StO2 (peripheral tissue oxygen saturation) after an ischemic - reperfusion event, rather than absolute value of StO2 are predictive of bad prognosis in a septic patient or in severe head injury. These variables are delta StO2, HbT (total tissue hemoglobin) and THI (muscular hemoglobin index). From these elements can be calculated others: RdecStO2 (StO2 decrease rate during ischemic event reflecting the tissue oxygen consumption) and RincStO2 (increase rate of StO2 during the reperfusion phase, corresponding to relative hyperemia). Both are expressed in %/s. The intensity of reperfusion can be quantified by delta THI.

A recent study in cardiac surgery did not show StO2 modifications under CBP but providing a small population sample and the type of monitoring used, this study might include some methodologic biais.

Finally, very recently a correlation between FMD and the variation of NIRS during a vascular occlusion test has been found in a young healthy subject, which only confirms the influence of endothelial function onto microcirculation.

C/ Syndecan - 1 concentration assessment: This molecule is closely linked to glycocalyx and is considered as the most predictive marker of its integrity.

Many factors might cause lesions of this structure (inflammation, hyperglycemia, CBP, etc). Its destruction causes further an impairement of endothelial function (vasoconstriction et vasodilation), mediated principally by endothelial factors.

Corticoids can be considered as the most effective molecule for glycocalyx protection.

Yet the investigators will analyze syndecan - 1 as a glycocalyx marker.

Material and methods :

After ethics commitee approbation and signed informed consent obtention, the investigators will enroll 60 patients scheduled for cardiac surgery with CBP (aorto - coronary bypass graft or valve replacement).

Patients randomized in the Corticoids group will receive 500mg of methylprednisolone in 100ml of NaCl 0,9% at anesthesia induction, Placebo group will receive 100ml of NaCl 0,9% - both solutions are prepared at the hospital pharmacy and only the pharmacist knows the preparation content (double blinding).

Exclusion criteria are age under 18, extreme emergency surgery if the testing would slow down the surgical treatment. The investigators also exclude the patients where the placement of pneumatic cuff would be impossible (amputation, AV fistula)

Anesthesia procedure:

All the patients receive the same anesthesia technique, based on propofol and remifentanil at objective of concentration. Relaxation is achieved by cisatracurium in continuous infusion. This type of anesthesia is the current practice in Erasme hospital. All the patients are equipped with radial artery catheter and central venous catheter, the use of pulmonary artery catheter is left to anesthetist's discretion.

Before starting the CBP, all patients receive heparin with target ACT (activated clotting time) above 480seconds. The priming of CBP consists on 500ml of gelatin (Geloplasma, Fresenius), with 500ml of cristaloids (Plasmalyte, Baxter) and 200ml of mannitol (Mannitol 15%, Baxter). Cardioplegy is achieved by blood infusion (T 36°) with potassium chloride. The CBP flow is continuous of 2.4L/m2 and mean arterial pressure is maintened between 60 and 90 mm Hg. Glycemia is controlled strictly from the moment of anesthesia induction by short acting insulin administrered continuously.

All patients are transferred to ICU after the surgery, sedated with propofol and remifentanyl until hemodynamic stability and absence of bleeding is achieved. Sedation is then stopped and patient is extubated.

Vascular occlusion test:

Arterial diameter variation and Doppler flow are measured at brachial artery above medial epicondyl, proximally to the pneumatic cuff placed at forearm. For echography analysis, a linear probe of 12 MHz will be used, held by a metallic support to allow more precision and better reproducibility. After initial diameter measurement, pneumatic cuff 50 mmHg will be inflated above the arterial pressure of the patient during 5minutes, in order to induce ischemia. Diameter variations in brachial artery are measured continuously from 30 secondes before releasing of the cuff until 3 minutes afterwards. FMD is evaluated by following parameters : baseline diameter and maximal FMD variation in pourcents. In numerous studies, FMD is calculated using the formula: FMD (%) = ([diametre of maximal dilation after cuff release - baseline diametre]/ baseline diametre) x 100. In order to reduce variability of the measrure, diameter variation will be adjusted to baseline diameter and to initiating stimulus assessed by Doppler flow. Data will be stocked and analyzed by Quipu FMD.

StO2 is measured by tissular spectrometer (Invos, Covidien) consisting of a photodetector, processor and a 15 mm optic probe. This proble is placed over thenar eminence of the arm without any IV line. StO2 is enregistered every 2 secondes.

Practically, after performing the initial echographic measurements, and after 3 minutes of stabilisation of StO2 (baseline), an ischemic event of 5 minutes is created by a fast pneumatic cuff inflation to a value of 50mmHg above patients systemic blood pressure. This will stop arterial and venous blood flow in the forearm, leading to a decrease of HbO2 (= rate of decrease StO2 (%/s)).

After 5 minutes the cuff will be rapidly deflated (<1seconds) inducing arterial diameter and flow increase - the variation is assessed by echography and Doppler. At the same time, StO2 increases (= StO2 increase rate (%/s)) and achieves higher values than baseline (= reactive hyperemia = delta diametre, flow velocity and StO2). The NIRS device measures also THi (tissular hemoglobin index) at baseline and 1 minute, its minimum et maximum.

The increase rate of arterial diameter reflects endothelial reactivity while StO2 increase rate is considered as a tissue = hemoglobin flow marker. Reactive hyperemia can be interpreted as a test of microcirculation reactivity. Rate of StO2 decrease/THi is considered as oxygen consumption marker.

Syndecan-1 analysis is performed on frozen plasma by ELISA technique, using specific antibodies (ELISA kit from Ray Biotech, Inc), this method has already been used in previous studies.

Timing of measurements:

Measures are performed on the day before surgery, at the ICU admission, at 24h and at the day 7 of hospital stay.

Glycocalyx function assessment:

The first analysis will be performed on the day before surgery, then after anesthesia induction, at ICU admission, at 24h and 48h.

Inclusion Criteria:

• All patients eligible for cardiac surgery

Exclusion Criteria:

• Age under 18years
• Allergy to corticoids
• Extreme emergency surgery
• Upper member vascular disease