Cerebral Flow-metabolism Coupling During Adult Surgery

The brain is a highly metabolic active organ, which receives approximately 15% of cardiac output with a normal cerebral blood flow (CBF) of approximately 50 mL/100 g/min. In normal circumstances, CBF is tightly coupled to local cerebral metabolism. An increase of cerebral metabolic activity of one region of the brain will trigger a corresponding increase of blood flow to that brain region. Conversely, a reduction of cerebral metabolic activity will lead to a reduction in blood flow. This tightly controlled physiological phenomenon is known as flow-metabolism coupling.

Flow-metabolism coupling is thought to be a feed-forward physiological protective mechanism wherein neuronal activity directly increases CBF, thereby matching the increased energy supply. In the event of cerebral hypoperfusion (i.e. reduction in CBF), brain vessels would be dilated in compensation and the local oxygen extraction fraction (OEF) would be increased. If cerebral hypoperfusion persists and is below the threshold, the activity of the brain cells would be suppressed, thereby resulting in the reduction of the cerebral metabolic rate (CMR) and cerebral metabolic requirement of oxygen (CMRO2). This flow-metabolism coupling mechanism is well studied using a PET scan in both healthy volunteers and patients with pathological conditions. Importantly, this physiological property of flow-metabolism in response to brain ischemia may be used as an earlier indicator of cerebral ischemia during surgery. In other words, monitoring that indicates a sudden drop in CMRO2 or increase in OEF may serve as an early marker of brain ischemia. However, due to technical limitations, bedside detection of such subtle physiological changes is challenging, especially in the intraoperative settings. In addition, many physiological, pathological, and pharmacological factors may affect this flow-metabolism coupling relationship and hence affect the interpretation of CMRO2/OEF changes as an indicator of brain ischemia. Additional studies are required to address how anesthetics may confound the interpretation.

Different anesthetic agents may alter flow-metabolism coupling in different ways. Cerebral metabolic rate is determined by a number of factors, including the functional state of the nervous system, anesthetic drugs and body temperature. The effect of individual anesthetic drugs on the CMR has been well studied. In short, most anesthetic drugs suppress the CMR in a dose-dependent manner, with the exception of ketamine and nitrous oxide (N2O). The anesthetic suppression of the CMR is mainly driven by the reduction of electrophysiologic activities. Previous studies have showed several anesthetics, including barbiturates, isoflurane, sevoflurane, desflurane, propofol, and etomidate can increase plasma concentrations and cause progressive suppression of EEG activity and a concomitant reduction in the CMR.

To further complicate the issue, during surgery, the anesthesiologist is required to maintain adequate cerebral perfusion pressure to ensure that there is sufficient CBF to meet metabolic demands. Induction of general anesthesia is often associated with a reduction in the mean arterial blood pressure(MAP) and cerebral perfusion pressure which is attributable to a decrease in cardiac output (CO) and systemic vascular resistance (SVR). Many intraoperative conditions such as blood loss and surgical positioning, may further compromise patient blood pressure and hence, CBF. Phenylephrine and ephedrine are commonly administered during neurosurgical procedures to counteract cardiovascular depression and treat anesthesia-related hypotension. Near-infrared spectroscopy (NIRS) studies in anesthetized patients suggest that phenylephrine reduces regional cerebral oxygen saturation (StO2) compared with ephedrine, despite a marked increase in the MAP. The results of these studies are conflicting and are various limitations.

NIRS has long been recognized as a promising neuromonitor due to the relative transparency of tissue to light in the near-infrared range and the dependency of light absorption on the oxygenation of haemoglobin, which provides a non-invasive method of measuring StO2. In addition, NIRS is extremely safe, inexpensive and provides real-time monitoring. Commercial systems are available and have been used for neuromonitoring during a variety of surgical procedures with an associated risk of brain injury, including coronary artery bypass and carotid endarterectomy.1,2 Despite the advantages of NIRS, its diagnostic accuracy for detecting intra-operative ischemia is controversial.3 A recent systematic review concluded that:

"NIRS has low sensitivity and high specificity to identify intra-operative ischaemia … Extracranial signal contribution was highly variable." 4

This conclusion reflects the two main limitations with commercial NIRS systems. First, the relationship between StO2 and ischemia is not direct since StO2 not only depends on CBF but also on the metabolic demand of the tissue. Second, light absorption in the scalp and skull is a major source of signal contamination. In fact, more than 80% of the NIRS signal originates from the extra-cerebral tissues since the distance from the skin surface to the brain ranges between 1.0 to 1.5 cm. As a result, hemodynamic changes in the superficial tissues can overshadow brain-related signals. Because of these two limitations, a specific StO2 threshold to identify intra-operative ischemia has not been established.

Our team has been developing technologies to tackle the two main limitations with commercial NIRS systems. With regards to enhancing depth sensitivity, the investigators have developed a time-resolved (tr)NIRS system that measures the time photons take to travel through tissue.5 Based on the principle that time equals distance, light that interrogates scalp and skull is recorded earlier than light that reaches the brain.

To provide continuous blood flow monitoring, the investigators have combined trNIRS with a complementary optical technology known as diffuse correlation spectroscopy (DCS). Instead of measuring light absorption, DCS measures temporal signal fluctuations caused by the movement of light scatterers, which in tissue is dominated by microvasculature blood flow (i.e. tissue perfusion).8 Like NIRS, DCS can be affected by signal changes in superficial tissue, although it has the intrinsic advantage that blood flow is considerably higher in the brain compared to the scalp (~ 5:1).9 To separate the two signal contributions, our system acquires data at two source-detector distances: a short distance to monitor scalp blood flow and a longer distance with greater sensitivity to CBF.10 The investigators have previously used DCS to monitor CBF stability in preterm infants during the first days of life and during cardiac surgery.11,12 A recent paper investigating the use of DCS to monitor CBF during carotid endarterectomy reported an average reduction in perfusion of 57% with clamping13.

Our hybrid trNIRS/DCS, which is shown has the ability to measure CBF and StO2 simultaneously. With this combination, oxygen metabolism in the brain, which is referred to as the cerebral metabolic rate of oxygen consumption (CMRO2), can be calculated by combining the two measurements.

Based on this new developed bed-side technique, the investigators plan to assess how CBF and CMRO2 change under general anesthesia and their responses toward phenylephrine and ephedrine during hypotension.

All study participants will be recruited and consented adhering to the local ethics guidelines. The attending anesthesiologists and the surgeons will not be blinded because of the nature of the intervention. However, patients and outcomes assessors will be blinded as to the treatment intervention. All patients will receive standard anesthesia care and the surgical and anesthetic management of the patients will be conducted in a standard fashion and will not be altered in this study. Anesthesia will be induced with fentanyl of 1-2 mcg/kg, propofol of 1.5-3 mg/kg, rocuronium of 0.6-0.9 mg/kg. Automatic blood pressure, electrocardiography, pulse oximetry, capnography, oesophageal temperature, and inspired and expired oxygen and CO2 concentration will be monitored. Controlled ventilation with 50% oxygen in air will be applied and the end-tidal CO2 levels will be maintained at 35-40 mmHg. The oxygen saturation is maintained above 96%. FiO2 can be increased if arterial desaturation occurs. Body temperature will be maintained above 36 ℃. After surgery, the patients will be extubated and transferred to the PACU. The depth of anesthesia will be continuously monitored either by bispectral index (BSI) or patient state index (PSI).

All patients will be monitored by the trNIRS/DCS device from anesthesia induction to completion of surgery. The NIRS sensor will be placed on the patient's forehead.

Before anesthesia induction, preoperative baseline cerebral hemodynamic and metabolic parameters such as ScO2, CBF, CMRO2, OEF will be measured for 1 minute. Monitoring will be continued during anesthesia induction and tracheal intubation. After anesthesia induction, post-induction cerebral hemodynamic and metabolic parameters will be obtained.

After tracheal intubation with airway secured, maintenance and vasopressor agents will be administrated according to the allocated group. The four possible allocated groups are as follows:

Group 1: Propofol-based anesthetic maintenance with phenylephrine used as the vasopressor Group 2: Propofol-based anesthetic maintenance with ephedrine used as the vasopressor Group 3: Sevoflurane-based anesthetic maintenance with phenylephrine used as the vasopressor Group 4: Sevoflurane-based anesthetic maintenance with ephedrine used as the vasopressor

Maintenance agents

• For the propofol group, the patient will receive propofol as their maintenance agent during surgery. The typical dose is 150-200 mg/kg/min.
• For the sevoflurane group, the patient will be maintained with 1 MAC of sevoflurane during surgery.
• Both propofol and sevoflurane are commonly used maintenance agents, and the depth of anesthesia will be kept at PSI of 30-40 (normal range 25-50) regardless of the choice of agents. Anesthesiologists use both agents frequently, and in our daily practice, the choice of anesthetic agents is mainly based on personal preference, with the exception of patients with specific contraindications such as malignant hyperthermia or drug allergy.

Vasopressor agents

• For the phenylephrine group, the patient will receive phenylephrine infusion as the primary vasopressor of choice.
• For the ephedrine group, the patient will receive ephedrine infusion as the primary vasopressor of choice.
• Vasopressor will only be administrated to maintain MAP within 0-20% above the baseline. In general anesthesia practice, ephedrine and phenylephrine can be administrated as a bolus injection or infusion. In this study, an infusion regime is used because of the necessity of maintaining a stable blood pressure during repeated measurements of cerebral hemodynamics in a steady state in each group. In patients with refractory hypotension despite either phenylephrine or ephedrine infusion, other vasopressor or inotropes will be used. For patients who did not receive any vasopressor (e.g. no hypotension during the entire surgery) or who receive multiple types of vasopressors, the investigators will analyze their data using both an intention-to-treat and as-per-protocol approach.

The neurological outcomes of patients in the post-anesthesia care unit (PACU) will also be collected via routine physical exam.