20 Intraoperative Neurophysiological Monitoring in Surgery for the Craniovertebral Junction

10.1055/b-0034-81397

20 Intraoperative Neurophysiological Monitoring in Surgery for the Craniovertebral Junction

Morota, Nobuhito

With the recent advances in neurosurgery and neuroimaging, a more aggressive approach to lesions that were once regarded as inoperative is now possible. The craniovertebral junction (CVJ) is one the most complex and challenging areas in skull base and spinal surgery. It is essential for neurosurgeons to avoid damage to critical neural and vascular structures while removing lesions in the CVJ. Intraoperative neurophysiological monitoring can contribute to safe surgical performance.

Intraoperative neurophysiological monitoring comprises evoked potential monitoring and functional mapping. Evoked potential monitoring involves the continuous observation of the functional integrity of a specific neural pathway. It enables direct feedback in terms of the integrity of neural function and informs the neurosurgeon as to what is going on during surgery. Functional mapping covers neurophysiological localization of a specific neural structure or pathway in the surgical field. It guides the neurosurgeon through the functional anatomy. Thus, the neurosurgeon can avoid direct damage to neural structures. To perform functional mapping, the neurosurgeon has to interrupt surgery from time to time to perform stimulation. This is the shortcoming of functional mapping. In general, evoked potential monitoring and functional mapping are combined. This is not always the case in CVJ surgery, as most of the surgical pathology is in the extradural space, and functional mapping cannot be applied unless cranial or peripheral nerves are involved. Evoked potential monitoring thus serves a more important role than functional mapping during this part of CVJ surgery.

In this chapter, the author describes clinical aspects of intraoperative neurophysiological monitoring in surgery for the CVJ that was performed mainly at the National Center for Child Health and Development (NCCHD) in Tokyo, Japan. Special emphasis is placed on clinical application of motor evoked potential (MEP) monitoring and its usefulness for surgery in and around the CVJ.1

Case Illustration

A 15-year-old boy diagnosed with Hajdu-Cheney syndrome had several previous surgeries at other institutions, including foramen magnum decompression and syringosubarachnoidal shunt. However, progressive gait disturbance, tetraparesis, and intractable headache forced him to live a bedridden life. Swallowing disturbances developed slowly, and he was referred to our institution. Preoperative computed tomography (CT) and magnetic resonance imaging (MRI) ( Fig. 20.1a ) revealed a marked basilar impression, with the clivus running horizontally and the tip of the odontoid process penetrating the brainstem. Because of the decompression of the suboccipital bone that was previously done, the cerebellar hemispheres hung over the upper cervical spine. Due to this condition, anterior decompression through the transoral transpalatal approach was chosen as a surgical treatment.

The question here was to identify the best intraoperative neurophysiological monitoring method that could be applied during the anterior decompression procedure performed in this critically ill young patient. Conventional evoked potential monitoring, such as the somatosensory evoked potential (SEP) or the auditory brainstem response (ABR), may be helpful in such cases to monitor the functional integrity of the brainstem and spinal cord. However, the neurosurgeon’s main concern during such surgery is that the motor function can be completely lost if further compression to the brainstem occurs. The answer in this case was MEP monitoring.2,3 Figure 20.2 shows pre- and postoperative recordings of the MEP, and Fig. 20.3 demonstrates continuous MEP monitoring during the transoral transpalatal anterior decompression. Although there was some reduction of the MEP amplitude by the end of the surgery, the MEP remained robust throughout the procedure, and the boy awoke with no further neurological deterioration. Postoperative CT and MRI ( Fig. 20.1b ) revealed satisfactory anterior decompression of the brainstem and spinal cord after surgery. Preoperatively, syringomyelia was present, and it decreased in size after surgery was performed. The boy received occipitocervical/upper cervical instrumentation 1 month later. He was discharged and was able to walk with the aid of a walker.

**Fig. 20.1 A 15-year-old boy with Hajdu-Cheney syndrome with severe basilar impression.** a Preoperative computed tomography (CT) and magnetic resonance imaging (MRI) demonstrate that the odontoid process seems to penetrate the brainstem. b Postoperative neuroimaging shows satisfactory brainstem decompression after surgery. The cervicothoracic syrinx also improved after surgery.

Fig. 20.2 Pre- and postoperative motor evoked potentials (MEPs) show approximately the same features of the muscle MEP recorded from the abductor pollicis brevis (APB) muscles. The results indicate that the motor function of the patient should be preserved during the decompression surgery.

**Fig. 20.3 Records of continuous MEP monitoring during the surgery are shown. Polarity of the transcranial stimulation was switched periodically, and the MEP was recorded from the contralateral APB (C3**+**/C4**−**indicates the left side stimulation, C4** +**/C3**−**the right side).**

Intraoperative Neurophysiological Monitoring: Motor Evoked Potentials

MEPs are elicited by transcranial electrical stimulation and recorded from either the spinal epidural space or the limb muscles. Electrodes for transcranial electrical stimulation are placed at C3 and C4 (10–20 international electroencephalographic [EEG] electrode system).4 The anode is a stimulating electrode, contrary to peripheral nerve stimulation. When unilateral stimulation is preferred, a C3/C4+ versus Fz− or C3/C4+ versus Cz− stimulating montage is used to elicit MEPs ( Fig. 20.4 ). Parameters for transcranial stimulation are given in Table 20.1 .

The MEP recorded from the spinal epidural space (epidural MEP) was obtained through the catheter electrode inserted percutaneously on the lower cervical or upper thoracic level ( Fig. 20.5 ). The spinal epidural MEP consists of D (direct) and I (indirect) waves. The D wave is elicited by activation of the cortical motoneurons directly, and the I wave from its transynaptical activation via the cortical interneurons.5 Single transcranial stimulation can evoke several I waves following the D wave. Therefore, because a D wave reflects specific information regarding the functional integrity of the corticospinal tract (CST) and is less influenced by anesthesia and other nonsurgical factors, it is used for intraoperative MEP monitoring.

The advantages of epidural MEP are that the D wave amplitude is usually stable, and its change correlates with the damage to the CST.6 Additionally, the D wave can be elicited by a single transcranial stimulus, and muscle relaxants can be used during surgery without influencing D wave monitoring. The disadvantages are that the placement of the recording electrode into the spinal epidural space is an invasive procedure, and evaluation of the individual CST is difficult.

**Fig. 20.4 Placement of electrodes for transcranial electrical stimulation for MEP monitoring based on the 10–20 international electroencephalographic (EEG) system.**

MEP recorded from the limb muscles (muscle MEP) is obtained from electrodes placed at the target muscle. The preferable muscle for muscle MEP monitoring is the abductor pollicis brevis (APB) for the upper extremity, and the tibialis anterior (TA) or the abductor hallucis brevis (AHB) for the lower extremity. Transcranially, a short train of stimuli is applied to elicit muscle MEP because during anesthesia, multiple descending drives are necessary to bring the resting potential of the α-motor neurons up to the firing level, which will consequently transmit signals to the peripheral nerves and muscles.7,8 Figure 20.6 shows an example of muscle MEP recorded from an 18-year-old male patient with Chiari I malformation. Muscle MEP was elicited by transcranially applying a short train of stimuli (C1/C2, train of five stimuli) and continuously recording from the bilateral APB, TA, and right diaphragm.

**Table 20.1** Parameters for transcranial stimulation for motor evoked potentials
Position of electrodes	C3/C4 or C1/C2
Waveform	Square wave
Train	N = 5
Interstimulus interval	2–4 msec
Duration of stimulation	0.5 msec
Frequency	4.3 Hz
Intensity	Suprathreshold intensity (< 200 mA)

The advantage of muscle MEP monitoring is that it enables surgeons to evaluate the individual motor function of each extremity without using an invasive procedure, such as placing an electrode in the spinal epidural space.3,9 The disadvantages are that the amplitude of muscle MEP is often unstable, and it fluctuates during surgery. In addition, muscle relaxants are not used with anesthesia, and total intravenous anesthesia is required for intra-operative monitoring of muscle MEP. It is necessary to confirm that all responses are present before starting the MEP monitoring by checking the train of four stimulation techniques being applied to the peripheral nerve. This is important to do so as to exclude any influence of residual muscle relaxant used for intubation purposes or that may be accidentally administered later.

Fig. 20.5 Difference between the muscle MEP and epidural MEP. The muscle MEP is evoked by a transcranial train of stimuli and recorded from limb muscles. The epidural MEP is evoked through a single transcranial stimulus and recorded from a percutaneously inserted electrode placed in the epidural space. In general, the recorded response from the epidural electrode consists of a D wave followed by one or more I waves.

Fig. 20.6 An 18-year-old male patient with Chiari I malformation. The MEP was recorded from the upper extremity (APBs), lower extremities (tibialis anterior muscles, TAs), and right diaphragm by transcranial train of stimuli. For this patient, electrodes for the transcranial stimulation were placed on the C1 (anode) and C2 (cathode).

Clinical Application of MEPs for CVJ Surgery

MEP monitoring is a practical intraoperative neurophysiological procedure for surgeries of the CVJ. Information obtained from MEP monitoring can be useful feedback on the well-being of the nerve structure at critical stages of the surgery.

The first step of CVJ surgery is the positioning of the patient’s head. Flexion or rotation of the neck is usually required for better access to the lesion. Manipulations of the neck under anesthesia can produce further compression of the spinal cord in some patients. MEP monitoring will help ensure that positioning of the patient on the operating table does not further compromise nerve structures. Figure 20.7 shows a complex atlantoaxial dislocation associated with os odontoideum. The hypo-plastic C1 lamina and superior left facet are invaginated into the posterior fossa. MEP is recorded before and after positioning of the patient on the operating table to confirm that the lesion does not produce further compression of the spinal cord at the CVJ ( Fig. 20.8 ).

Fig. 20.7 Preoperative MRI and CT scans of a 4-year-old girl with atlantoaxial dislocation due to os odontoideum. The cervicomedullary junction was severely compressed. Details of the bony anomaly were revealed by the three-dimensional (3D) CT scan. Part of the hypoplastic C1 lamina was invaginating into the foramen magnum, contributing to the cervicomedullary compression. 1, os odontoideum; 2, base of the odontoid process; 3, part of the hypoplastic right C1 lamina; 4, dislocated left C1 superior facet.

Fig. 20.8 Intraoperative position of the patient (upper left). A bone flap harvested from the skull used for the occipital-C2 posterior fixation (lower left). MEP monitoring shows no significant change in amplitude before and after positioning, and the MEP remains the same amplitude after surgery (right).

Fig. 20.9 On the left: 3D and sagittal reconstruction of CT scans. Atlantoaxial subluxation was corrected, and no stenosis at the foramen magnum was observed. On the right: records of continuous MEP monitoring during the surgery demonstrate stability of response.

MEPs were continuously monitored during the occipital-C2 fixation ( Fig. 20.9 ). When passing a thread at the foramen magnum or C2 lamina, the surgeon must be sure that the procedure is safe. Continuous MEP monitoring during surgery can help avoid neural damage.

Finally, MEP monitoring helps confirm that surgery was successful in terms of preserving the motor function even before the patient wakes up from anesthesia. In general, if the spinal MEP amplitude remains > 50% from the baseline value, the patient’s motor function will be the same as that seen preoperatively. MEP amplitude < 50% indicates that the patient has suffered serious motor deterioration.9,10 Regarding muscle MEP monitoring, the presence of muscle MEP indicates that motor function was preserved. The loss of muscle MEP amplitude does not necessarily suggest that motor function is lost if at the same time spinal MEP is present. This indicates that there is simply transient postoperative motor deterioration that will recover later. For a more precise prediction of motor outcome, it has been strongly recommended that both spinal and muscle MEPs be monitored.11

Case Illustration

A 1-year-old baby girl diagnosed with achondroplasia was referred to the NCCHD because of developmental delay in motor function. Her parents noticed marked snoring during sleep with the head extended. MRI and CT revealed prominent stenosis at the foramen magnum and spinal cord compression. Before positioning the patient on the operating table, the MEP was recorded, and she was turned prone with the neck kept in a neutral position. The MEP showed no apparent deterioration after positioning ( Fig. 20.10 ). The MEP was continuously monitored during surgery. When the foramen magnum decompression was started, the MEP amplitude suddenly augmented, then significantly deteriorated ( Fig. 20.11 ). It was speculated that the partial decompression caused focal spinal cord compression at the edge where the decompressed dura bulged. Foramen magnum decompression was quickly accomplished. Consequently, the muscle MEP amplitude increased with lower amplitude, when compared with the baseline at the end of foramen magnum decompression. The amplitude of muscle MEP gradually recovered thereafter until the end of surgery. The patient showed no neurological deterioration. Her locomotion improved after surgery, and she was able to walk after 1 month.