The study consist of evaluation in cardiac surgery with cardiopulmonary bypass (CPB) setting the ability of PCO2 derived variables (ΔPCO2, ΔPCO2/C(a-v)O2 ratio), compared to lactate and ScVO2 to predict major postoperative adverse events.

Cardiac surgery with cardiopulmonary bypass is associated with serious morbidity and mortality especially in moderate and high-risk patients. This procedure is associated with systemic inflammatory response as a consequence of cardiopulmonary bypass, surgical insult and genetic background of patients leading to organ injury and worse outcome. This pitfall may be worsened by hemodynamic changes with inadequate hemodynamic management.

During and after CPB, substantial changes in macrocirculation and microcirculation are observed and sustain impairment may result in reduced oxygen delivery and/or impaired oxygen extraction. The main consequence is cellular dysorexia that may trigger postoperative organ dysfunction. Rapid identification of cellular dysorexia and rapid hemodynamic management are therefore among key strategies that may reduce mortality.

In this purpose various marker can be considered. Traditionally lactatemia is considered as surrogate of anaerobic metabolism resulting from ischemia. However it interpretation may be challenging particularly in case of reduced hepatic clearance, use of epinephrine or massive blood transfusion. Venous or central venous oxygenation (S(c)VO2), a surrogate of oxygen extraction that is believed to reflect balance between oxygen delivery and consumption, is considered as an acceptable alternative as it was shown to be associated with organ dysfunction in various clinical setting. Nevertheless ScVO2 suffers from the difficulties to define adequate threshold as very high S(c)VO2 as well as low S(c)VO2 may be associated with poor outcome. Recently PCO2 derived dysorexia and perfusion markers have been shown to be predicting outcome in both septic patient and high risk surgical patient. Central venous to arterial difference in PCO2 (ΔPCO2) a global perfusion index is show to be correlated to microcirculation dysfunction and may reflect impaired tissue perfusion. In high risk non-cardiac surgical patients and in septic patient, ΔPCO2 predicted worse outcome better than S(c)VO2 and lactate. Besides this performance may be improve when using a clinically available surrogate based on ΔPCO2. When anaerobic metabolism occurred as a result of sustained hypoxia, CO2 production increases therefore the respiratory quotient (CO2 production (VCO2) and oxygen consumption (VO2) ratio) increases. VCO2/VO2 can be simplified as ΔPCO2 /Ca-vO2 ratio, where Ca-vO2 is arteriovenous O2 content difference. All these variables have never been compared in cardiac surgery setting and their association with microcirculation impaired is poorly documented. The hypotheses is that ΔPCO2, and ΔPCO2 /Ca-vO2 ratio may better predict major postoperative adverse events than blood lactate and S(c)VO2 after cardiac surgery with CPB.

Inclusion Criteria:

• 18 years old or more
• Cardiac surgery with cardiopulmonary bypass
• Tip of a central venous catheter positioned in superior vena cava or right atria
• Arterial catheter correctly positioned

Exclusion Criteria:

• KDIGO 3 AKI prior to surgery
• Hepatic insufficiency prior to surgery
• Extracorporeal life support prior to surgery '
• Live expectancy lower than 48 hours
• pregnancy

Patients undergoing cardiac surgery with cardiopulmonary bypass in the university hospital of Lille.