Neuromonitoring During Cardiac Surgery Utilizing Cardiopulmonary Bypass Procedures

Background

• In 2006, an estimated 7 million inpatient cardiovascular operations and procedures were performed in the US alone of which 448,000 involved coronary artery bypass graft (CABG) surgery.

• Cardiac surgery utilizing cardiopulmonary bypass procedures (CPB) consists of a cascade of dynamic surgical and anesthetic events. The incidence rates of early postoperative neurological and/or neuropsychological complications have ranged from 0-100% with the most commonly reported ranging between 35-50%. In particular, impaired cognitive function can persist in up to 35% of patients for 12 months.

• A recent postoperative study by Newman and colleagues found that 42% of bypass patients still experienced cognitive decline five years later. Both the rEEG and qEEG have been used to "provide a sensitive measure of synaptic activity within the cortical mantel". Although, intraoperative neuromonitoring of the rEEG during CPB is nearly as old as extracorporeal circulation, brain monitoring has not yet become a routine tool in the repertoire of surgical monitoring for these procedures. However, is should be noted that in certain studies the use of intraoperative rEEG and qEEG has significantly lowered neurological deficits, shortened postoperative recovery, and reduced hospital costs.

Anesthetic Effects on the rEEG During Cardiac Surgery

The use of rapid-acting synthetic opioids (such as fentanyl and sufentanil) during cardiac surgery is a well-established practice for cardiac anesthesia. For rEEG, anesthetic induction begins with the appearance of diffuse theta and some delta which is maximal frontally. Within 1 to 2 minutes following the emergence of an irregular bifrontal delta, global and more synchronous, monomorphic, delta activity may prevail, depending upon the dose and the individual. Over the next 2-5 minutes the global rEEG pattern evolves and stabilizes into a more polymorphic or irregular, slow-wave activity, comprised mostly of delta waves [65, 66].

Although the level of anesthesia may be maintained in a range that permits a stable baseline for the rEEG, alterations due to ischemia are easily masked by high-dose anesthesia and thus may go undetected. One method for detecting ischemic changes during this type of anesthetic regime, which is typically even more complicated by induced-hypothermia during CPB, involves the use of relative brain power in the delta frequency bandwidth (see Section 6.3). Lastly, it should be noted that many centers currently use anesthetic protocols combing inhalational and intravenous agents similar to those used for CEA surgery.

rEEG and qEEG Changes Associated with Cardiac Surgery

Like neuromonitoring for CEA surgery, changes in the rEEG and qEEG have been used for the detection of cerebral ischemia, objective administration of anesthesia, blood pressure control, and "cerebral protection" during cardiac surgery. Increasing neural activity (i. e., augmentation of higher frequency components or decreased lower frequency activity in the analog signal, and increased mean and median frequencies, and a higher SEF95 index) which is non-pathologic, probably signifies decreasing anesthetic effects.

Conversely in the presence of stable anesthesia, slowing of the rEEG with a progressive loss of amplitude and frequency content, burst-suppression, and even electrocortical silence (isoelectric waveforms) are indications of pathological synaptic depression.

Although, the rEEG can be well-suited for the detection and correction of anesthetic imbalance or intraoperative seizure activity, electroencephalographic depression caused by pathologic factors suchas ischemia or hypoxia may not be easily discriminated from non-pathologic influences such as hypothermia or deepening anesthesia (e. g., induced-hypothermia with onset of bypass or a bolus of high-dose narcotics, respectively).

As stated by Edmonds et al. [10], one major function of electroencephalographically-based interventions using rEGG and qEEG during cardiac surgery is to optimally match perfusion with metabolic demand during the critical periods of surgery. Possible conditions or times when this relationship is compromised are listed below.

1) Marked slowing of the rEEG associated with cannulation and onset of bypass related to the low-oxygen, priming volume of the bypass

machine. The magnitude and duration of the rEEG changes are directly related to the priming volume and inversely related to the size of the patient. Thus, smaller patients are more likely to exhibit marked slowing of the rEEG at cannulation.

2) Release of the aortic cross-clamp with its attendant transient hypotension is often coupled with slowing of the rEEG that may persist until effective pulsatile perfusion is re-established.

3) Depression of the rEEG may also be expected at the completion of rapid rewarming. Because of the demand of a hypermetabolic state induced by 39o- 40o C, blood outpaces the delivery of the mechanical or recovery of the cardiac pump. This may be viewed as a ‘cerebral anginal attack.’

Lastly, another important use for monitoring rEEG and qEEG during cardiac surgery is to provide objective maintenance and documentation of burst-suppression and isoelectric patterns during deep hypothermia and/or barbiturate protection prior to the initiation of circulatory arrest [68, 69]

Alarm Criteria for Ischemic Thresholds Using qEEG During Cardiac Surgery

The benefits of neural monitoring using qEEG for the detection of cerebral ischemia have been discussed by various authors. As a matter of historical interest, two studies will be reviewed that utilized commerically-available software packages which represented a pioneering effort for neuromonitoring using qEEG during cardiac surgeries involving CPB. In particular,Aromet al. demonstrated that patients who were “brain-monitored” with interventional criteria implemented as indicated below, presented postoperatively with only 5% new, global, neurological deficits as compared to an incidence of 40% in patients “brain-monitored” without interventions based on neural events. A power drop index (PDI) was calculated and compared to baseline measures. The PDI is a numerical indicator of the severity and duration of the decrease from baseline power level for each channel. The greater the decrease in power and/or the longer it lasts, the higher the PDI. The PDI would accumulate when the power level dropped to less than 40% of baseline level for a given channel.

The criteria for intervention CPB were:

• -Any drop in power to 25% of baseline activity during CBP

• -Any asymmetry or lateralized drop in power during pump CBP

The methods of intervention were:

• -Increase cerebral perfusion (Mean Arterial Pressure - Central Venous Pressure = Cerebral Perfusion Pressure; MAP - CVP = CPP) to 60 - 65 mmHg

• -Increase CPB pump flow

• -Increase MAP using vasopressor (in this case, neo-synephrine)

• -Readjust the venous cannula

• -Increase blood CO2

• -Readjust the arterial cannula for a lateralized deficit

Another study also employed a proprietary software package to evaluate cerebral ischemia during cardiac surgery using CPB. In particular, a drop in the relative low-frequency power (1.5-3.5 Hz) was chosen as a single, quantitative, electroencephalographic descriptor using Cerebrovascular Intraoperative Monitor (CIMON; Cadwell Laboratories, Inc., Kennewick, WA)). Relative delta power appeared to be insensitive to moderate changes in body temperature (deliberate-induced hypothermia) and level of opioid anesthesia, but was a statistically significant indicator of cortical dysfunction when using standard deviations in z-scores from an individualized reference or selfnormative data. Despite the advantages of EEG monitoring in some studies, other studies have found no relation between rEEG changes and neurophysiologic outcome after cardiac surgery due to the wide variability in the rEEG seen during these procedures.

Although displays of the qEEG and derived indices can be very useful for the detection of cerebral ischemia as summarized above, great care and caution should be paid to its exclusive use. qEEG is subject to a variety of unpredictable influences of sampling error, environmental, and statistical artifact. High-dose narcotics and hypothemic suppression of the activity of the rEEG can produce artifactually-induced increases in relative delta, theta, and beta activity. In addition, during ischemic suppression of the rEEG, the SEFmay remain unchanged in instanceswhere imperceptibly small amounts of high frequency artifact contaminate the signal.