CENTRAL nervous system (CNS) injury is a potential complication of many surgical procedures. Intraoperative monitoring of CNS function frequently is used to warn of possible injury. Modalities include local or regional anesthesia with the patient awake, a wake-up test during general anesthesia, electroencephalography, electromyography, and various evoked potentials (EPs). Evoked potentials involving spinal cord pathways are particularly useful during procedures on or near the spinal cord and may allow localization of possible injury because of the topography of the monitored tracts. Intraoperative monitoring of motor EPs in humans was first reported in 1983 and has since gained popularity as an improved method of detecting possible intraoperative spinal cord injury. [4-6]Initial descriptions of the transcranial technique noted a lack of adverse effects, and it has more recently been described as “free of complication.”We report the first incidence of a significant complication of monitoring transcervical motor EPs.
A 14-yr-old girl presented for anterior and posterior spinal fusion for correction of kyphoscoliosis secondary to mild spondyloepiphyseal dysplasia. She had mild asthma, which was well controlled with intermittent inhaled albuterol, but was otherwise healthy. Her only other medication was oral iron; she had no allergies. She weighed 58 kg, with a heart rate of 84 beats/min and a blood pressure of 100/78 mmHg. Examination was remarkable only for kyphoscoliosis. Hematocrit concentration was 33.6%.
Intravenous induction with incremental total doses of 2 mg midazolam, 500 [micro sign]g fentanyl, 100 mg lidocaine, 250 mg thiopental, and 6 mg pancuronium was uneventful. Nasotracheal intubation was accomplished without difficulty, and arterial and central venous access were established. Anesthesia was maintained with 0.5-0.7% inhaled enflurane, 2-3 [micro sign]g [middle dot] kg-1[middle dot] h-1intravenous fentanyl, and bolus intravenous pancuronium. Inhaled nitrous oxide was added at final surgical closure.
Bilateral upper and lower extremity somatosensory EPs via percutaneous needle electrode stimulation of the ulnar nerve at the wrist and of the posterior tibial nerve at the ankle were recorded at the popliteal fossa, the posterior cervical spine, and three scalp sites. Stimulation frequency was 4.7 Hz, intensity was 25 mA, and duration was 0.3 ms; 300 responses were averaged per recording. Bilateral lower extremity neurogenic motor EPs via percutaneous needle electrode stimulation at the first and fourth cervical vertebral levels were recorded at the popliteal fossa. Stimulation frequency was 4.7 Hz, intensity was 400 V, and duration was 0.3 ms; 100 responses were averaged per recording. Somatosensory and motor EPs were recorded in an alternating fashion, with 5-15 min between modalities. The electrodes used to stimulate for motor EPs (The Electrode Store; Buckley, WA) were 7 cm in length, insulated to within 3 mm of the tip, and placed percutaneously onto the vertebral lamina. Electrodes were placed by a technician before incision and taped to the skin without bending or other manipulation. Both modalities provided reproducible data indicating intact sensory and motor pathways; a representative recording of the motor EP is shown in Figure 1.
Anterior T10-L4 spinal fusion was accomplished uneventfully in the right lateral decubitus position, after which the patient was turned prone for the posterior T4-sacrum procedure. Central venous pressure remained unchanged at 9 mmHg after this change in position, and the patient's intravascular volume was thought to be adequate. To this point, she lost 600 ml of blood, received 2,200 ml of isotonic crystalloid, and made 0.25 ml [middle dot] kg-1[middle dot] h-1of clear, light-colored urine. No disturbance of the EP electrodes was noted. After the change in position, however, multiple transient episodes of profound hypotension without change in heart rate were noted, unrelated to surgical or anesthetic manipulation. These episodes occurred exclusively during stimulation for motor EPs. Hypotension developed within seconds of stimulation, the blood pressure consistently falling from approximately 100/60 mmHg to as low as 40/20 mmHg, with no change in heart rate. During two separate episodes, administration of 100 [micro sign]g intravenous phenylephrine after approximately 20 s produced a prompt response, the blood pressure rising to 115/60 mmHg and the heart rate decreasing from 90 to 80 beats/min; these gradually returned to baseline over several minutes. Without intervention, hypotension invariably resolved within approximately 60 s after stimulation for motor EPs ceased. No episode of hypotension lasted for more than 2 min; all EPs remained intact throughout.
The transcervical needle electrodes used to stimulate motor EPs were removed and replaced with similar electrodes at the seventh cervical vertebral level. Additional electrodes were placed by the surgeon during direct visualization at high thoracic vertebral levels within the surgical field. Stimulation via both sets of new electrodes yielded satisfactory motor EPs and elicited no hypotension.
Initial posterior vertebral distraction caused a significant reduction in the amplitude of the right posterior tibial somatosensory EP, without significant change in any other EP. The distraction was modified, intravenous dopamine 5 [micro sign]g [middle dot] kg-1[middle dot] min (-1) was begun to maintain a mean arterial blood pressure of at least 70 mmHg, and high-dose intravenous methylprednisolone (30 mg/kg over 1 h, followed by 5.4 mg [middle dot] kg-1[middle dot] h-1for 23 h)was begun at the surgeon's request. All EPs gradually returned to baseline levels. The patient was extubated without incident at the end of the procedure, demonstrated normal neurologic function throughout, and was eventually discharged with no apparent neurologic deficits.
Intraoperative monitoring of CNS function has evolved considerably during the past several decades. The wake-up test, first described in 1973, may allow detection of intraoperative neurologic injury during procedures requiring general anesthesia, [11,12]but entails the risks of intraoperative recall, pain, and patient injury. Somatosensory EPs, first used in the late 1970s, offer continuous intraoperative monitoring of CNS function, but with regard to the spinal cord provide information only about the ascending sensory tracts of the dorsal columns. Furthermore, numerous anesthetic agents influence signal quality. Motor EPs, by contrast, offer continuous intraoperative monitoring of the descending motor tracts of the anterior spinal cord, an area particularly prone to ischemic damage. First described for human use in 1983, motor EPs have been reported to improve detection of intraoperative spinal cord injury ; some motor EPs have been reported to be more resistant than somatosensory EPs to the effects of various anesthetic agents. [5,14]As with any monitoring modality, the ability of motor EPs to predict ultimate neurologic deficit is imperfect. 
When, as in this case, motor EPs are monitored using needle electrodes, possible minor complications include pain, bleeding, infection, and damage to local structures. However, the small size of the needles used, their insertion after induction, and their removal before emergence render such complications infrequent. Profound hypotension, as in this case, represents a potentially serious complication. Failure to recognize the cause might prompt unnecessary and possibly detrimental changes in surgical or anesthetic technique. Repetitive episodes could also predispose structures at risk, particularly the anterior spinal cord, to ischemic injury.
We suspect that the profound transient hypotension observed in this patient represents an electrically induced sympathectomy, analogous to that which can be induced traumatically in acute spinal shock or pharmacologically in total spinal anesthesia. Specifically, we suspect that the transcervical needle electrodes used to stimulate motor EPs became malpositioned during the transition from the right lateral decubitus to the prone position. Subsequent stimulation of inhibitory efferent sympathetic pathways, or of afferent pathways mimicking a state of high sympathetic tone, might induce a reduction in sympathetic output, causing vasodilatation, hypotension, and lack of reflex tachycardia. At cessation of the stimulus, sympathetic tone would quickly return to baseline. Hypotension after electrical stimulation of motor cortex in animals was well described by the 1940s. More recent animal studies have described similar hypotension after electrical stimulation of frontal cortex and have even suggested sympathectomy as a possible mechanism. We suspect that the hemodynamic changes observed in this patient represent a spinal analogue of these phenomena.
When faced with this complication, the anesthesiologist must first rule out other causes of intraoperative hypotension. If monitoring motor EPs appears to be the cause, the modality may be discontinued, with resultant potential loss of ability to detect intraoperative CNS injury. Alternatively, needle electrodes may be replaced with a magnetic coil, [19,20]if this modality is available. Finally, transcervical needle electrodes may be repositioned or replaced with electrodes over the scalp or within the surgical field. 
Intraoperative monitoring of CNS integrity is of significant potential benefit in many surgical procedures. Transcervical motor EPs are a promising addition to such monitoring, but are not without at least one potentially serious complication.