Background: Different anaesthetic agents have been shown to have different protective effects upon heart, brain and renal function under ischaemic conditions (oxygen starvation). Cardiopulmonary bypass takes over the work of the heart and the lungs during heart surgery, but oxygenation of vital organs such as the brain and heart may not be perfect, and can produce brain or heart damage as a consequence. Propofol and desflurane are commonly used anaesthetic agents, and there has been recent research to suggest that anaesthetic agents may provide some protection during periods where inadequate oxygenation occurs, with the potential to reduce the degree of organ damage. Both types of anaesthetics are used for cardiac surgery with anaesthetists choosing between them largely on the basis of personal preference.

Aim: To determine whether the use of either propofol or desflurane as the primary anaesthetic agent, can lead to differences in postoperative brain function, total morbidity or cost, following coronary artery surgery with cardiopulmonary bypass.

Methods: Patients will be recruited by professional research staff and will be randomised into one of two groups (90 in each group). They will receive a standardized technique for anaesthesia, cardiopulmonary bypass and postoperative ICU treatment. The only difference between the 2 groups will be as to which anaesthetic agent they receive during the surgical period, desflurane or propofol. Measurements will involve i) brain function testing before and 3 months after surgery ( a set of 10 verbal or manual tests), ii) incidence of delirium in the immediate postoperative period (a survey form), iii) incidence of total postoperative morbidity and iv) cost of hospital stay. Data collection will be by anaesthesia and research staff and a neuropsychologist will employed for performing the brain function testing.

Anticipated timeline: Initial recruitment completed by 15-18 months following trial commencement. Follow up completed 3 month after the last enrolment. Data validation, statistical analysis and manuscript preparation completed by 24 months.

Aims To investigate whether there are differences in postoperative cognitive function, total morbidity or cost, following cardiac surgery determined by the use of either propofol or desflurane as the primary anaesthetic agent.

Hypothesis

The null hypothesis is that there are no differences in postoperative cognitive function, total morbidity or economic cost based on the type of primary anaesthetic delivered.

Participants: Following human ethics committee approval, and informed written consent, 180 patients undergoing elective cardiac surgery will be randomized to receive desflurane or propofol as the primary anaesthetic agent during surgery. This will be conducted at the Royal Melbourne Hospital. Data will be collected preoperatively, intraoperatively and throughout the entire hospital stay, and including a three-month follow-up.

Eligibility criteria

Inclusion Criteria

Adult male and female patients aged 18 years or older, undergoing on-pump elective coronary artery bypass surgery with general anaesthesia.

1. Off-pump cardiac surgery
2. Require surgery for acute coronary syndrome
3. Dialysis dependent renal dysfunction
4. Severe liver dysfunction as determined by liver transaminases 1.5X greater than normal.
5. Pre-existing diagnosis of schizophrenia, dementia recent stroke or cognitive disorder
6. Recent alcohol/drug abuse/intoxication
7. Re-do Coronary Artery Grafts
8. Coronary Artery Grafts plus other surgery

Interventions

Patients will receive either propofol or desflurane as the primary anaesthetic agent for the duration of this surgery. The following minor differences between groups are outlined:

1. For anaesthesia induction, the desflurane group will receive inhalational sevoflurane, as desflurane is pungent and irritant to the airway making inhalational induction of anaesthesia difficult. Following induction of anaesthesia with sevoflurane, the anaesthetic will be changed to desflurane at 0.5-2 MAC as necessary (this will include desflurane administration during cardiopulmonary bypass via the oxygen inlet to the circuit).
2. Patients in the propofol group will have general anaesthesia induced and maintained by propofol (effect site steering using an Asena TCI pump and levels maintained from 1.5-3µg/ml using the Marsh pharmacokinetic profile). No volatile anaesthesic agent will be allowed in this group including during CPB.

The following anaesthesia, surgery, cardiopulmonary bypass, and sedation techniques will be common for both groups:

1. Anaesthesia co-induction / sedation with fentanyl (2-5 ug/kg) and midazolam (0.025- 0.05mg/kg). Following induction of anaesthesia, patients will be maintained on a combined intravenous infusion of fentanyl (1.5 ug/kg/hr) and midazolam (0.025-0.05 mg/kg/hr) with Fi02=1.
2. Anaesthesia monitoring: In addition to specified routine anaesthetic monitoring as per Australian and New Zealand College of Anaesthetists guidelines for general anaesthesia (38), all patients will have intra-arterial pressure monitoring, pulmonary artery (PA) pressure monitoring, nasopharyngeal temperature monitoring and transoesophageal echocardiography. These are standard monitoring for patients undergoing cardiac surgery at the Royal Melbourne Hospital.
3. Hemodynamic management: metaraminol or nitro-glycerin can be administered to control systemic arterial blood pressure within a systolic arterial pressure range of 100-140 mmHg. Beta blockade can be used at the anaesthetists discretion for either haemodynamic control, or management of ventricular arrhythmias.
4. Cardiopulmonary bypass: the venous reservoir of the cardiopulmonary bypass circuit will be primed with 2L of crystalloid solution (Plasma-lyte(R) Baxter Healthcare) and maintained at a temperature of 35 degrees Celsius. Cardiopulmonary bypass will be performed in a standardized technique with cannulation of the ascending aorta and with a single two-staged right atrial-caval cannula with non-assisted venous drainage. After initiation of cardiopulmonary bypass the patient's nasopharyngeal temperature will be allowed to drift progressively down to 34 degrees Celsius at which point it will be maintained by a heat exchanger until rewarming is initiated. Patients' haemoglobin will be maintained above 70 g/l. The ascending aorta will be cross-clamped and cardiac arrest will be induced by administration of tepid blood cardioplegia at a temperature ranging from 20-25 degrees Celsius. The ratio of blood to the initial crystalloid cardioplegia required to achieve arrest will be 4:1 in order to obtain a potassium concentration in the induction cardioplegia of at least 20 mmol/l. Arrest will be achieved by antegrade administration of cardioplegia through an aortic root catheter, followed by retrograde administration through a coronary sinus catheter. Left ventricular distension will be monitored for by transoesophageal echocardiography. Further doses of maintenance cardioplegia are given following completion of graft anastomoses. Rewarming of the patient will commence at the beginning of the last distal anastomosis (or 15 minutes prior to aortic cross clamp removal). The temperature of the heat exchanger used to achieve this will not exceed 37 degrees Celsius. Atrial pacing wires will be attached if electrical activity has not been established after 5 minutes of reperfusion or if pre-operative heart rate is less than 60 per minute or if the patient is on B-adrenergic antagonists. The patient will remain on full cardiopulmonary bypass support for a period of time approximately equal to 20-25% of cross clamp time or until the anaesthetist, perfusionist, and surgeon are satisfied that cardiac function is appropriate to achieve separation.
5. Post CPB hemodynamic management: Cardiac output will be measured by a thermodilution technique. Dobutamine can be administered for low cardiac output (cardiac index less than 2.0 L/min/m2) and noradrenaline can be used for low SVR syndrome if felt appropriate by anaesthetist, surgeon, or intensivist concerned.
6. Postoperative sedation: following skin closure, all anaesthesia agents will be ceased. A low dose propofol infusion will be permitted for postoperative ICU sedation until they are suitable for extubation.The aim is for a Ramsay sedation score of 2 upon initial return to ICU then once haemodynamically stable, to aim for a score of 4. Patients will be escorted to the intensive care unit and will be ventilated until they have achieved accepted standard criteria for extubation (awake and co-operative, warm, haemodynamically stable, PaCO2<50mmHg, PaO2> 100 on FiO2 = 0.4). Patients will be allowed to awaken in ICU and extubated.
7. Rescue protocol: In both groups additional rescue analgesia will be allowed in the form of fentanyl, or morphine, paracetamol, tramadol, or indomethacin as necessary to provide adequate analgesia.

Endpoints

Primary Endpoint - neurological protection 1. Neurocognitive function testing (preoperative, prior to hospital discharge, and at 3 months postoperation)

Secondary Endpoints

1. Delirium as assessed by confusion assessment method (CAM).
2. In-hospital morbidity as determined by composite morbidity score

Tertiary endpoints

1. Cost of post-operative care

The following endpoints are cumulative and reported for the entire hospital stay:

Composite morbidity/mortality score This will be a weighted morbidity score. It will score 3 points for death, 2 points for a major morbid outcome and 1 point for any minor outcome. Major morbid outcome is defined as any outcome that is permanent (e.g. stroke) or has a natural history in that it could lead to death, but without actually causing death. It is derived from morbidity likely to result from inadequate tissue perfusion or immunomodulation.

Economics Hospital economics will be assessed on multiple criteria;

1. Time spent in ICU
2. Time spent in hospital
3. Cost of all blood products
4. Re-operation
5. Pathology and other investigations needed while in hospital
6. Cost of drugs administered

Cost/economics estimations are taken from Australian published studies and index to 2006 prices. Investigation tests are based on rebates from the Medical benefit schedule. Both are listed below. The cost of re-operation is based on theatre time for four hours indexed to 2006 prices, and do not include doctors fees. The number of hospital days which will be charged as the hospital length of stay - ICU time.

The actual cost of the operation are "fixed costs" (this only relates to the primary operation and not to reoperations). The cost that we are looking at, therefore are potentially variable costs which may be impacted upon by the possible use of volatile anaesthesia, or propofol. Obviously a return to OR would involve costs involving OR time, disposables including CPB circuits, and drugs (including aprotinin and recombinant factor VIIa).

Sample size Estimates of power and minimal group size were obtained by performing a forward looking power analysis based upon the ability to show a significant difference in neurocognitive testing. This is based on our previous study showing a 38% incidence of neurocognitive abnormality in a group consistent with the propofol arm, and aiming for an effect size reduction of 50%. For an 80% power to reject the null hypothesis of no significance (P<0.05) in each group, the sample size is 90 patients in each group.

Randomization The treating anaesthetist will allocate randomization by the sealed envelopes method.

Implementation The patients will be recruited by professional research staff. Randomization will be implemented by the treating anaesthetist.

Blinding The study will be open label for the treating anaesthetist who will also collect intraoperative data, and patient (as they will either have an inhalational induction or an intravenous induction, blinding is not possible), Preoperative and postoperative data will be collected by professional research staff who will be blinded to the treatment protocol.

Statistical methods Continuous data collected over repeated measurement intervals will be examined by repeated measures ANOVA and adjusted for multisample asphericity by applying the Greenhouse Geisser correction. Categorical data will be analysed using Fischer's Exact test. Data will be corrected for multiple comparisons within families of endpoints using the Ryan-Holm Bonferroni correction. Intention to treat analysis will be performed.

A blinded interim analysis of results will be performed by the data safety monitoring committee upon conclusion of enrolment of 80 patients. Stopping values to terminate the study will be set at a P value < 0.001 for primary endpoints.

Timeline We anticipate initial recruitment will be complete 15-18 months following trial commencement. Follow up will be complete 3 month after the last enrollment. Data validation, statistical analysis and manuscript preparation will be complete by 24 months.

Inclusion Criteria:

• Coronary artery bypass surgery

Exclusion Criteria:

• Off-pump cardiac surgery
• Require surgery for acute coronary syndrome
• Dialysis dependent renal dysfunction
• Severe liver dysfunction as determined by liver transaminases 1.5X greater than normal.
• Pre-existing diagnosis of schizophrenia, dementia recent stroke or cognitive disorder
• Recent alcohol/drug abuse/intoxication
• Re-do Coronary Artery Grafts
• Coronary Artery Grafts plus other surgery