A prospective comparison of neurogenic mixed evoked potential stimulation methods: utility of epidural elicitation during posterior spinal surgery.

STUDY DESIGN: A prospective study of 50 patients who underwent posterior thoracic or thoracolumbar instrumented spinal surgery from June 1998 through June 1999. OBJECTIVES: To highlight the advantages and disadvantages of neurogenic mixed evoked potential (NMEP) data obtained with three methods of stimulation: epidural, spinous process, and percutaneous. SUMMARY OF BACKGROUND DATA: Reports in the literature have established the efficacy of epidural, spinous process, and percutaneous stimulation of the NMEP response. The three methods have not been compared for reliability, sensitivity, and specificity. METHODS: The study group consisted of 50 patients who underwent posterior thoracic or thoracolumbar instrumented spinal surgery from June 1998 through June 1999. Somatosensory-evoked potentials were used to monitor upper and lower extremities. An attempt to obtain NMEPs was made in all patients by using percutaneous (PERC-NMEP), spinous process (SP-NMEP), and epidural (EPI-NMEP) stimulation. These data were evaluated for reliability, sensitivity, and specificity. The number of minutes monitored in the postcorrection period were calculated for each method, and stimulus intensities were noted. RESULTS: In the current study, PERC-NMEPs were obtained in 88% of the patients and were maintained in 91% of those cases, SP-NMEPs were obtained in 96% and maintained in 77%, and EPI-NMEPs were obtained in 100% and maintained in 88%. Data collection continued in the postcorrection period for 46 minutes for PERC-NMEPs, 19 minutes for SP-NMEPs, and 23 minutes for EPI-NMEPs. The study group had no true-positive or false-negative findings. CONCLUSIONS: Results showed that EPI-NMEPs provide reliable data in a greater number of patients than either SP-NMEPs or PERC-NMEPs. However, PERC-NMEP data are readily maintained during and after the critical time window after corrective spinal maneuvers. The NMEPs elicited with both percutaneous and epidural stimulation have a useful role in an intraoperative spinal cord-monitoring protocol.

Efficacy of intraoperative monitoring for pediatric patients with spinal cord pathology undergoing spinal deformity surgery.

STUDY DESIGN: A retrospective study of 38 pediatric patients with spinal cord pathology who underwent corrective spinal deformity surgery from January 1989 through June 1998. OBJECTIVES: To report reliability and specificity in obtaining intraoperative data in this population. These data were compared with monitoring results obtained in a group of pediatric patients with idiopathic scoliosis. SUMMARY OF BACKGROUND DATA: Reports in the literature suggest intraoperative monitoring for patients with spinal cord pathology may be of limited value. No optimal monitoring protocol has been suggested for this population. METHODS: The study group consisted of 38 pediatric patients with a diagnosis of spinal cord pathology who underwent corrective spinal deformity surgery from January 1989 through June 1998. All patients had lower extremity function. Somatosensory and neurogenic motor evoked potentials were used to monitor neurologic status during surgery. These data were compared with data obtained in 429 pediatric patients with idiopathic scoliosis. Study patients were divided into Group I, those who had had spinal cord surgery (n = 20), and Group II, those who had not (n = 18). RESULTS: Somatosensory evoked potentials were obtained in 93.2% and remained consistent with baselines in 87.2% of the study group patients. Neurogenic motor evoked potentials were obtained in 50.8% of the study subjects and remained consistent in 76.6% of those cases. The false-positive rate was 27.1% in the study group, compared with 1.4% in the group with idiopathic scoliosis. The study group had no true-positive or false-negative findings. Group I data differed from Group II data. CONCLUSIONS: Intraoperative monitoring should be used in patients with spinal cord pathology who undergo surgery for spinal deformity. Monitoring should not miss a neurologic deficit but demonstrates greater variability, resulting in more frequent use of an intraoperative wake-up test.

Spinal cord monitoring: current state of the art.

The intraoperative application of evoked potential and electromyographic (EMG) monitoring has increased significantly over the last 2 decades. Cranial nerve monitoring is widely accepted and used by otologists, neurologic surgeons, and ophthalmologists. Direct and indirect techniques for assessing the peripheral nervous system are used by plastic and orthopedic surgeons when performing intraoperative nerve grafting. Myriad techniques and applications for monitoring the spinal cord and peripheral nervous system have been developed, evaluated, and used by orthopedic and neurologic surgeons involved in spinal surgery.

Somatosensory- and motor-evoked potential monitoring without a wake-up test during idiopathic scoliosis surgery. An accepted standard of care.

STUDY DESIGN: This was a retrospective study of 500 patients undergoing corrective surgery between 1987 and 1997 for spinal deformity caused by idiopathic scoliosis. OBJECTIVES: To report the sensitivity and specificity of somatosensory-evoked and neurogenic motor-evoked potentials monitoring and the requirements for an intraoperative wake-up test for all idiopathic scoliosis surgeries at a single institution. SUMMARY OF BACKGROUND DATA: Intraoperative monitoring is recommended for use during corrective spinal surgery. Accepted monitoring standards and requirements for an intraoperative wake-up test are still debated. METHODS: The study group consisted of 500 patients undergoing corrective surgery for idiopathic scoliosis between 1987 and 1997. All patients were monitored using somatosensory-evoked and neurogenic motor-evoked potential techniques, using a standard protocol developed at this institution. RESULTS: The false-positive rate (significant data change without postoperative neurologic deficit) was 0.014% (n = 7). The true-positive rate (degradation of data that met warning criteria, with a corresponding postoperative neurologic deficit) was 0.004% (n = 2). No false-negative results (normal data during with a postoperative neurologic deficit) were seen. The sensitivity of combined somatosensory-evoked and neurogenic motor-evoked potential data in predicting neurologic status was 98.6%, and the specificity of normal data predicting normal findings in a neurologic examination was 100%. CONCLUSION: Combined somatosensory-evoked and neurogenic motor-evoked potentials monitoring during idiopathic scoliosis surgery represents a standard of care that obviates the need for an intraoperative wake-up test when reliable data are obtained and maintained.

Neurogenic motor evoked potentials: a prospective comparison of stimulation methods in spinal deformity surgery.

Neurogenic motor evoked potentials (NMEPs) elicited by spinal cord stimulation via the spinous processes (SP-NMEP) have been widely accepted as a sensitive method of monitoring motor tract function. SP-NMEP requires additional surgical dissection as well as electrodes within the wound, making the method somewhat inconvenient. A less invasive percutaneous method of spinal cord stimulation (PERC-NMEP) has more recently been described. We prospectively compared the SP-NMEP and PERC-NMEP methods in 184 patients undergoing 225 surgical procedures. Although SP-NMEP responses were more readily obtainable than PERC-NMEP, the reliability of the two methods was not significantly different. Both methods were found to be sensitive to neurologic deficit. The present study suggests that when responses are obtained, the percutaneous method is reliable enough to obviate the spinous process method of monitoring the motor function of the spinal cord.

Somatosensory evoked potentials fail to diagnose thoracic outlet syndrome.

Somatosensory evoked potentials (SEPs) are used in the diagnosis of thoracic outlet syndrome (TOS), even as an indication for surgery. The purpose of this study was to evaluate the use of SEPs in the diagnosis of TOS. Twenty-one patients (mean age, 37 years) with TOS and 23 control subjects (mean age, 34 years) were included. Somatosensory evoked potentials of median and ulnar nerves were measured bilaterally in patients in both a relaxed and arms-elevated provocative position. A three-way analysis of variance showed no significant difference between the interpeak latencies of the TOS and control groups (p = .352). Significant differences were found in testing positions (p = .0014) and nerve tested (p = .001) in both groups. Therefore, this study suggests that SEPs are not helpful in the diagnosis of TOS.

Validity and reliability of spinal cord monitoring in neuromuscular spinal deformity surgery.

Somatosensory evoked potentials (SSEPs) have become a standard of care in surgery for spinal deformity. Recent reports in the literature have suggested SSEP monitoring is not efficacious in surgeries for patients with neuromuscular disease. Electrophysiologic data were retrospectively analyzed from 74 patients with various neuromuscular disorders undergoing spinal-deformity surgery from 1989 through February 1995 at this medical center. The monitoring protocol included SSEPs recorded from multiple sites located cortically, subcortically, and peripherally. Neurogenic motor evoked potentials (NMEPs) were also employed. Anesthetic regimens were controlled for compatibility with evoked potential monitoring. Use of this intraoperative monitoring protocol resulted in reliable data for 95% of the patient population having baseline responses. Findings suggest that evoked potentials can be used effectively during surgery for neuromuscular spinal deformity. Use of a specific protocol allowed acquisition of reliable data intraoperatively, suggesting these methods are a valid means of monitoring neurologic status.

Triggered electromyographic threshold for accuracy of pedicle screw placement. An animal model and clinical correlation.

STUDY DESIGN: This study consisted of a laboratory investigation of triggered electromyographic stimulation of pedicle screws placed in a pig spine, with a correlative prospective clinical series of lumbosacral pedicle screws stimulated in a similar fashion. OBJECTIVES: To determine the threshold of stimulus intensity necessary to confirm accuracy of lumbar pedicle screw placement via a triggered electromyographic peripheral response. SUMMARY OF BACKGROUND DATA: Documentation of lumbar pedicle screw placement is imperative to perform proper spinal instrumentation and to avoid perioperative complications. Previous electrophysiologic techniques using stimulation of a pedicle opening or pedicle screw with peripheral recording of electromyographic activity from the lower extremity muscles have been used to identify varying threshold values that indicate a break in the bony pedicle wall. METHODS: Six adult pigs had 107 pedicle screws placed bilaterally into the pedicles of the lumbar spine. These screws were stimulated with an ascending stimulus intensity until a peripheral triggered electromyographic response was recorded. Pedicle screws were placed in the pig either entirely in the pedicle (Group A), medial to the pedicle without direct contact to the nerve root and dura (Group B), or purposely medial to the pedicle with direct contact to the nerve root and dura (Group C). A correlative clinical series of 233 pedicle screws placed in 54 patients had a similar intraoperative neurophysiologic technique. RESULTS: In the animal model, the mean threshold differences were: Group A screws 21.9 mA, Group B screws 8.5 mA, and Group C screws 4.2 mA (P < 0.05). Ninety-three percent of the clinical Group A screws had threshold stimuli less than 8.0 mA, whereas Groups B and C screws had a mean threshold of 3.3 mA. CONCLUSIONS: Triggered electromyographic stimulation is a valuable aid to determine appropriate placement of pedicle screws. We recommend the following interpretation of threshold stimulus intensity: > 8 mA--screw entirely in the pedicle; 4.0-8.0 mA--potential for pedicle wall defect; < 4.0 mA--strong likelihood of pedicle wall defect with potential for nerve root and dura contact.

The clinical application of neurogenic motor evoked potentials to monitor spinal cord function during surgery.

The purpose of this study was to report results from 300 cases (177 children, 123 adults) administered somatosensory and neurogenic motor evoked potentials during surgery. Of these 300 cases, there were 16 cases of spinal fractures, 16 neurosurgical cases, 28 vascular cases, and 240 cases of elective posterior spinal deformity requiring instrumentation. Results indicated that somatosensory evoked potentials, especially cortical components, demonstrated greater variability than neurogenic motor evoked potentials. Variability was attributed to anesthesia and unknown factors. Neurogenic motor evoked potentials proved to be a more valid indicator of postoperative motor status than somatosensory evoked potentials. Based on their anatomic substrates and results from this study, it was recommended that somatosensory evoked potentials and neurogenic motor evoked potentials be used to monitor spinal cord function during surgery that would place that structure at risk.

Clinical correlation between degenerative spine disease and dermatomal somatosensory-evoked potentials in humans.

The relationship between clinical status and preoperative and intraoperative dermatomal somatosensory-evoked potentials was investigated as a function of test site. Results indicated that the specificity and sensitivity of dermatomal somatosensory-evoked potentials varied with level of involvement: L5 demonstrated greatest agreement with equal correlation at L3, L4, and S1. Variables that influenced correlation included history of previous surgeries, type of anesthetic used, interpretation criteria, and whether the patient was awake or asleep. It was concluded that dermatomal somatosensory-evoked potentials can provide the surgeon with diagnostic and intraoperative information regarding the functional integrity of single nerve root function.