Techniques

Many intraoperative techniques are adapted from common outpatient testing: EEG, brainstem auditory evoked potential (BAEP), and somatosensory evoked potential (SEP) tests, for example. EEG is used for surgery that risks cortical ischemia, such as aneurysm clipping or carotid endarterectomy. BAEP is used for procedures around the eighth nerve or when the brainstem is at risk in posterior fossa procedures. SEP is widely used for many kinds of procedures in which the spinal cord, brainstem, or sensorimotor cortex is at risk.

Other techniques are more specific to the operating room. Transcranial electrical motor evoked potential (MEP) tests are evoked by several-hundred-volt electrical pulses delivered to motor cortex through the intact skull. Recordings are from extremity muscles. This monitors the corticospinal tracts during cerebral, brainstem, or spinal surgery. Electrocorticography (ECoG) measures EEG directly from the exposed cortex. This guides the resection to include physiologically dysfunctional or epileptogenic areas while sparing relatively normal cortex. Direct cortical stimulation applies very localized electrical pulses to cortex through a handheld wand. The electricity disrupts cortical function such as language, which can be tested in patients awake during portions of the craniotomy. Stimulation near motor cortex can produce movement. These techniques identify language or motor regions so they can be spared during resections. Similar direct nerve stimulation is used for cranial and peripheral nerves to locate them amid pathological tissue and check whether they still are intact. One version is stimulation at the floor of the fourth ventricle or during brainstem resection to identify tracts and nuclei of interest. For spinal procedures using pedicle screws, risk is incurred to the nerve roots or spinal cord during screw placement. To reduce that risk, EMG is monitored while electrical stimulation is delivered to the hole drilled in the spine or the screw as it is being placed. If the hole or screw errantly has broken through bone into the spinal or nerve root canal, stimulation will elicit an EMG warning of misplacement. In-depth descriptions of each procedure is beyond the scope of this chapter. The reader is referred elsewhere for extensive coverage of intraoperative neurophysiological techniques (Nuwer, 2008).

Interpretation

Interpretation of intraoperative neurophysiology includes two categories. One is monitoring, in which baseline findings are established and subsequent findings are compared to baseline. Alarm criteria are set in advance based on knowledge of how much change is acceptable without risk. The other category, testing, is decision making about structures and limits of resection. Testing is used in several ways. One is to identify a structure, such as finding the facial nerve within pathological tissue where it may be difficult to identify. Another is to identify motor or language cortex prior to a resection. A third example is identifying which cauda equina root is L5, which is S1, and which is S2, or which is the sensory or the motor portion of a root.

Response to Change

The monitoring team quickly assesses whether a change is likely due to a technical, systemic, or surgical cause. Occasional transient significant changes occur without significant risk for postoperative neurological problems. Transient changes for a few minutes can occur without substantial risk of postoperative problems, especially if the neurophysiological findings return shortly to baseline. Risk of neurological complications is higher when changes remain through the end of the case. For example, a very high-risk situation is complete and permanent loss of EPs that previously had been normal and easily detected.

The surgeon reviews actions of the preceding 15 minutes that may have caused change. Surgical problems causing neurophysiological changes include direct trauma, excessive traction, excessive compression, stretching from spinal distraction, vascular insufficiency from compression, clamping, embolus or thrombus, and other clinical circumstances. Clamping a carotid during an endarterectomy commonly produces EEG changes within 15 seconds. Most other changes are cumulative over many minutes, leading to changes at many minutes after the offending action. Two factors compound that delay. Evoked potential recordings take 1 to several minutes to average and sometimes longer when electrocautery or other electrical noise is ongoing.

Many surgical or anesthetic actions can be taken. Remedial measures can be implemented depending on the recent surgical actions. The surgical maneuver underway can be paused, stopped, or reversed. Resection can be halted. A graft or fixation instrumentation can be removed or repositioned. If no recovery of EPs occurs within 20 minutes, the patient can be awakened on the operating table and told to move the legs (“wake-up test”). Blood pressure can be increased. Steroids sometimes are given, although the literature about their usefulness is controversial. A vascular shunt can be placed, clamped vessels can be unclamped, a clip can be adjusted, or transected aortic intercostal arteries can be reimplanted. Retractors can be repositioned. Spine distraction can be reduced. Causes can be sought through inspection and exploration for mechanical or hematoma nervous system impingement. Motor and language identified can be avoided during resection. Systemic or local hypothermia or barbiturate-induced coma can be implemented for nervous system protection. Lowering of cerebrospinal fluid pressure by free drainage can be used in some cases of spinal ischemia. Hemoglobin level can be increased by transfusion. Other interventions also are used.

Prediction of Deficits

Intraoperative monitoring is effective at preventing many postoperative neurological complications. Risks depend on severity and duration of IOM changes. Transient changes that revert to baseline within a few minutes are rarely accompanied by postoperative deficits. On many occasions, these represent clinically significant problems that are identified and corrected promptly and completely—the goal of monitoring. In other cases, transient changes are false alarms. Both are combined in outcome studies as “false-positive” monitoring events, since their causes cannot be directly separated. Outcomes studies show false positives in several percent of cases. Persistent changes of moderate degree are accompanied by a risk of new neurological postoperative impairment in about half of cases (Nuwer et al., 1995). Sometimes such postoperative neurological impairment is less than might have occurred if monitoring had not initiated interventions that partially corrected the problem. Severe monitoring changes often are accompanied by postoperative neurological deficits. Some are due to intraoperative problems that were identified promptly but could not be completely corrected.

Anesthesia

Many inhalation anesthetics substantially affect cortical function (Sloan and Heyer, 2002). Agents commonly used attenuate or abolish cortical EP recordings. Limiting the inhalation anesthetic dose often produces satisfactory anesthesia compatible with monitoring. Boluses of centrally active medication are discouraged because they can cause transient IOM changes. Continuous-drip medication delivery is preferred. Much less susceptible to anesthetic effects are the nonsynaptic pathways such as peripheral nerve conduction techniques. Subcortical monosynaptic pathways are less affected than cortical polysynaptic pathways. For example, in SEP monitoring, brainstem peaks remain relatively robust despite inhalation anesthesia levels that nearly eliminate cortical peaks in the same pathway. MEPs tolerate inhalation anesthesia poorly, so most MEPs are carried out using total intravenous anesthesia with propofol, a centrally excitatory anesthetic agent, as opposed to the more inhibitory gas inhalation agents. Turning this effect around, anesthetic and drug effects can be monitored by the degree of evoked potential or EEG changes. When a barbiturate-induced cortically protective burst suppression state is desired, EEG is the primary tool to identify that sufficient depth has been achieved.

A surgical patient’s core temperature may drop 1°C or more. Limb temperature may drop more. Axonal conduction velocity depends on is temperature, so peak latencies increase as temperature drops. Monitoring can help identify therapeutic temperature effects. When a hypothermia-induced cortically protective isoelectric state is desired, EEG is the primary tool to identify that sufficient depth has been achieved.

Clinical Settings

Box 32E.1 lists many clinical conditions and types of surgery for which IOM is used. Intracranial posterior fossa cases commonly use BAEP, SEP, and cranial nerve EMG monitoring. Typical applications are cerebellopontine angle and skull base tumor resection, brainstem vascular malformation and tumor resection, and microvascular decompressions (Møller, 1996). Intracranial supratentorial procedures include resections for epilepsy, tumors, and vascular malformations as well as for aneurysm clipping. These use a combination of EEG and SEP monitoring together with functional cortical localization with direct cortical stimulation and ECoG. Surgery of the carotid, aorta, or heart may use EEG to monitor hemispheric function or assess the need for shunting or adequacy of protective hypothermia (Plestis et al., 1997). Some also use or prefer SEPs for these vascular cases.

Spinal surgery is the most common setting for IOM. Disorders include cervical diskectomy and fusion for myelopathy, stabilization for deformities such as scoliosis, resection of spinal column or cord tumors, and stabilization of fractures. Both SEP and MEP often are used to assess the posterior columns and corticospinal tract functions. The use of MEP depends on the case, since it requires total intravenous anesthetic and incurs some movements during surgery. As a result, some spinal cases still are done with SEP alone. In cases involving pedicle screw placement, EMG is monitored to detect screw misplacement (Shi et al., 2003). Spinal cord monitoring also is used for cardiothoracic procedures of the aorta that jeopardize spinal perfusion (Jacobs et al., 2006). Peripheral nerve monitoring is carried out for cases risking injury to the nerves, plexus, or roots. Testing also can determine which segments of a nerve are damaged when performing a nerve graft.

Outcomes have been assessed most thoroughly for spinal cord surgery. In one large multicenter study of SEP IOM in 100,000 cases of spinal surgery, the rate of false-positive alarms was about 1%. The rate of false-negative cases was about 0.1%, which were those cases with postoperative neurological deficits in which monitoring did not raise an alarm. Some were minor transient changes, and others were neurological deficits that started during the hours or days postoperatively. The rate of major intraoperative changes missed by SEP monitoring was 0.063%. The risk of paraplegia was 60% less among the monitored cases when compared to historical and contemporaneous controls. That amounted to 1 case out of every 200 that did not have paraplegia when monitoring was used (Nuwer et al., 1995). To improve even further on these SEP IOM monitoring outcomes results, MEPs have been used more recently together with SEP for many spinal procedures. The expectation is that the rate of false-negative cases and postoperative neurological deficits will be reduced even further.

Summary

Intraoperative monitoring is used routinely to monitor the nervous system during procedures that present significant risk of brain, spinal cord, or nerve injury. Intraoperative testing is used to identify nerves and eloquent cortex and to define regions and structures for resection. Many different intraoperative procedures are now available, and they are applied in a wide variety of surgical procedures. Studies show significant improvement in patient outcome with monitoring.