The effect of optimizing cerebral tissue oxygen saturation on markers of neurological injury during coronary artery bypass graft surgery

Abstract

Surgical revascularization of the coronary arteries is a cornerstone of cardiothoracic surgery. The enduring nature of coronary artery bypass grafting (CABG) bespeaks of its history and proven efficacy. However, cerebral deoxygenation during on-pump coronary artery bypass graft surgery may be associated with adverse neurological sequelae. Advanced age and the incidence of preoperative co-morbidity in patients presenting for coronary artery bypass graft surgery increases the potential for stroke and other adverse perioperative outcomes (Murkin, Adams, Quantz, Bainbridge and Novick, 2007). It is hypothesized, that by using the brain as an index organ, interventions to improve cerebral oxygenation would have systemic benefits for cardiac surgical patients. In an attempt to predict those patients that are predisposed to cerebral complications, investigators have used neurological monitoring ie, Near infrared spectroscopy (NIRS) to enhance detection of hypoxic conditions associated with neurological injury (Hoffman, 2006). Serum S100B protein has been used as a biochemical marker of brain injury during cardiac surgery. Elevated levels serve as a potential marker of brain cell damage and adverse neurological outcomes (Einav, Itshayek, Kark, Ovadia, Weiniger and Shoshan, 2008). Aims and Objectives of the study This prospective, quantitative, interventional study was carried out to maintain cerebral tissue oxygen saturation during cardiopulmonary bypass above 75% of the baseline level by implementation of a proposed interventional protocol. The analysis of S100B which is a marker of neurological injury and optimization of regional cerebral oxygen saturation would allow for the formulation of an algorithm which could be implemented during on-pump coronary artery bypass graft surgery as a preventive clinical measure further reducing the risk of neurological injury. Central venous lines (CVP) are inserted routinely during cardiac surgery. Central venous oxygen saturation is a global marker of tissue oxygenation. A secondary aim of the study was to determine if a correlation existed between central venous and cerebral tissue oxygen saturations. If a positive correlation existed then central venous oxygen saturation could be used as a surrogate measure of cerebral tissue oxygen saturation during on-pump coronary artery bypass graft surgery. This study is one of the first done in the South African population group. Methods Forty (40) patients undergoing on-pump coronary artery bypass graft surgery were recruited at Inkosi Albert Luthuli Central Hospital. Patients were randomized into a control group (n=20) and interventional group (n=20) using a sealed envelope system. The envelope contained designation to either group. Envelopes were randomly chosen. Intraoperative regional cerebral oxygen saturation (rSO2 ) monitoring with active display and treatment intervention protocol was administered for the interventional group. In the control group regional cerebral oxygen saturation monitoring was not visible to the perfusionist operating the heart lung machine during cardiopulmonary bypass (blinded). Recording of regional cerebral saturation was conducted by an independent person (another perfusionist) who was not involved in the management of the case so as to ensure that no interventions were carried out on the control group. Arterial blood samples for the measurement of serum S100B were taken pre and postoperatively. An enzyme immunoassay (ELISA) was used for the quantitative and comparative measurement of human S100B concentrations for both groups. Central venous oxygen saturation was monitored from the CVP using the Edwards Vigileo monitor. Cerebral monitoring constituted the use of Near infrared spectroscopy monitoring using the Invos 5100c, Somonetics Corp, Troy MI monitor. Adhesive optode pads were be placed over each fronto- temporal area for cerebral oxygen measurement. During cardiopulmonary bypass, eight time period measurements of mean arterial pressure (MAP), heart rate, temperature, activated clotting time (ACT), patient oxygen saturation (SpO2), partial pressure of carbon dioxide (pCO2), haematocrit, lactate, pH, haemoglobin (Hb), base excess (BE), potassium (K+), sodium (Na+), glucose, calcium (Ca2+), central venous oxygen saturation (ScvO2), cerebral tissue oxygen saturation (rSO2), fraction inspired oxygen (FiO2 ), sweep rate, pump flow rate (cardiac index), and percentage isoflurane per patient were taken. The time periods when data was recorded included: 5 minutes after onset of cardiopulmonary bypass, aortic cross clamping, after cardioplegic arrest, during distal anastomosis, during proximal anastomosis, during rewarming, after aortic cross clamp release and before termination of cardiopulmonary bypass. Baseline measurements were also taken. Clinical data recorded for both groups included: the number of grafts performed, cardiopulmonary bypass time, cross clamp time, red blood cells administered (packed cells), amount of adrenalin infused and total cerebral desaturation time. A prioritized intraoperative management protocol to maintain rSO2 values above 75% of the baseline threshold during cardiopulmonary bypass was followed. Cerebral desaturation was defined as a decrease in saturation values below 70% of baseline for more than one minute. Interventions commenced within 15 seconds of decrease below 75% of baseline value. Results The results of the study show that there was a highly significant difference in the change in S100B concentrations pre and post surgery between the interventional and control groups. The intervention vii group showed a smaller increase in S100B concentration of 37.3 picograms per millilitre (pg/ml) while the control group showed a larger increase of 139.3 pg/ml. Therefore, the control group showed a significantly higher increase in S100B concentration over time than the intervention group (p < 0.001). Maximizing pump flow rates was the most common intervention used (45 times) followed by maintaining partial pressure of carbon dioxide to approximately 40 mmHg (28 times), increasing mean arterial pressure by administration of adrenalin (11 times) and administration of red blood cells to increase haematocrit (11 times). There was a highly statistically significant treatment effect within the intervention group for each of the above interventions compared with no intervention. The above mentioned interventions significantly affected right and left cerebral oxygen saturations. However, administration of red blood cells was not found to significantly increase right (p = 0.165) and left (p = 0.169) cerebral oxygen saturation within the intervention group. The study highlighted a significant difference between the intervention and control groups in terms of cerebral desaturation time (p <0.001). The mean desaturation time for the control group was 63.85 minutes as compared to 24.7 minutes in the interventional group. Cerebral desaturation occurred predominantly during aortic cross clamping, distal anastomosis of coronary arteries and aortic cross clamp release. Predictors of cerebral oxygen desaturation included, partial pressure of carbon dioxide (pCO2), temperature, pump flow rate (LMP), mean arterial pressure (MAP), haematocrit, heart rate (HR) and patient oxygen saturation (SpO2). Central venous oxygen saturation was not significantly related to right (p = 0.244) or left (p = 0.613) cerebral oxygen saturations. Therefore central venous oxygen saturation cannot be used as a surrogate measure of cerebral tissue oxygen saturation during on-pump coronary artery bypass graft surgery. viii Conclusion These findings demonstrate the positive effect of optimizing cerebral oxygen saturation using an interventional protocol on markers of neurological injury (S100B). Optimization of pump flow rate, partial pressure of carbon dioxide and mean arterial pressure would result in increased cerebral oxygen saturation levels and a reduction in neurological injury. Therefore, an algorithm incorporating these interventions can be formulated. Monitoring specifically for brain oxygen saturation together with an effective treatment protocol to deal with cerebral desaturation during on-pump CABG must be advocated.