Integrity Of Spinal Cord Research Articles

Magnetic motor evoked potential monitoring in the rat.

The authors conducted a study to provide an objective electrophysiological assessment of descending motor pathways in rats, which may become a means for predicting outcome in spinal cord injury research. Transcranial magnetic motor evoked potentials (TMMEPs) were recorded under various conditions in awake, nonanesthetized, restrained rats. Normative data were collected to determine the reproducibility of the model and to evaluate the effect of changing the stimulus intensity on the evoked signals. In addition, an experiment was performed to determine if the TMMEPs produced were the result of auditory startle response (ASR) potentials elicited by the sound generated by the movement of the copper coil inside its casing during magnetic stimulation. Transcranial magnetic motor evoked potentials were elicited after magnetic stimulation. At 100% stimulus intensity, the mean forelimb onset latency was 4.2 +/- 0.39 msec, and the amplitude was 9.16 +/- 3.44 mV. The hindlimb onset latency was 6.5 +/- 0.47 msec, and the amplitude was 11.47 +/- 5.25 mV. As the stimulus intensity was decreased, the TMMEP onset latency increased and the response amplitude decreased. The ASR potentials were shown to have longer latencies, smaller amplitudes, and were more variable than those of the TMMEPs. These experiments demonstrate that TMMEPs can be recorded in awake, nonanesthetized rats. The evoked signals were easy to elicit and reproduce. This paper introduces noninvasive TMMEPs as a new technique for monitoring the physiological integrity of the rat spinal cord.

New monitors of neurological functions — part 2

This article is the second of two parts. In this section, techniques for monitoring cerebral oxygenation, cerebral biochemistry, intracranial elastance and spinal cord integrity are described. The developments of certain of the new monitors including near-infrared spectroscopy, brain tissue oxygen tension measurement, intracerebral microdialysis, monitors of compliance and techniques for measuring motor evoked potential are discussed in detail. New applications for some established monitors (jugular bulb venous oximetry, intracranial pressure waveform analysis) are also described.

Adjunctive Therapy for Spinal Cord Protection During Thoracoabdomina Aortic Aneurysm Repair

Paraplegia, resulting from spinal cord ischemia during thoracoabdominal aortic aneurysm (TAA) repair, continues to be a devastating complication. The incidence of neurological complications, including paraplegia and paraparesis following TAA repair, ranges from 4% to 32% and averages 13% for nondissecting TAA and higher for dissecting TAA. Our current understanding of spinal cord ischemia associated with TAA repair has evolved from animal research and clinical experience. The pathophysiology of spinal cord ischemia is intricately related to three physiological variables. These include the severity and duration of spinal cord ischemia, neuronal reperfusion after reestablishment of spinal cord blood flow, and the neuronal metabolic rate during the ischemic insult. We have developed a multimodality approach to the prevention of neurological deficits, during and after TAA repair, which includes minimizing the severity of spinal cord ischemia, reducing the anticipated reperfusion phenomenon, and lowering the spinal cord metabolic rate. Over the past 16 years, the senior author has undertaken surgical repair of 265 TAAs using a multimodality approach in the protection of spinal cord integrity. In our experience, a combination of adjunctive therapies is critical to minimize the ischemic interval, reduce the neuronal reperfusion injury, and decrease spinal cord metabolism. These techniques have evolved over time, resulting in an overall neurological deficit rate of 4.5% and a neurological deficit at the time of hospital discharge of 1.9%. This article will outline our multimodality approach for spinal cord protection during TAA repair.

A comparison of ECG cancellation techniques applied to the surface recording of somatosensory evoked potentials

The use of somatosensory evoked potentials (SEPs) for determining spinal cord integrity in a clinical setting is well known. The poor signal-to-noise (SNR) ratio of such measurements has led to ensemble averaging being employed to extract the signal from the background noise. Over the thoracic region, the poor SNR is largely the result of interference generated by the cardiac musculature. Therefore, any reduction in the level of this cardiac interference will greatly improve the performance of any SEP monitoring system. Three methods were investigated as techniques to reduce the cardiac interference in SEP measurements. These were clipping, gating and adaptive noise cancellation (ANC). It was found that, although clipping and gating performed as well as, if not better than, ANC, these techniques were both very sensitive to the setting of a threshold level. The linear ANC scheme employed circumvented the problem associated with a threshold level, and with an order of 50, SNR improvement figures of approximately sixfold were achievable over the entire thoracic region.

Intraoperative monitoring of motor evoked potentials: a review of 116 cases.

We reviewed the results of motor evoked potential (MEP) and somatosensory evoked potential (SEP) monitoring during 116 operations on the spine or spinal cord. We monitored MEPs by electrically stimulating the spinal cord and recording compound muscle action potentials from lower extremity muscles and monitored SEPs by stimulating posterior tibial or peroneal nerves and recording both cortical and subcortical evoked potentials. We maintained anesthesia with an N2O/O2/opioid technique supplemented with a halogenated inhalational agent and maintained partial neuromuscular blockade using a vecuronium infusion. Both MEPs and SEPs could be recorded in 99 cases (85%). Neither MEPs nor SEPs were recorded in eight patients, all of whom had preexisting severe myelopathies. Only SEPs could be recorded in two patients, and only MEPs were obtained in seven cases. Deterioration of evoked potentials occurred during nine operations (8%). In eight cases, both SEPs and MEPs deteriorated; in one case, only MEPs deteriorated. In four cases, the changes in the monitored signals led to major alterations in the surgery. We believe that optimal monitoring during spinal surgery requires recording both SEPs and MEPs. This provides independent verification of spinal cord integrity using two parallel but independent systems, and also allows detection of the occasional insults that selectively affect either motor or sensory systems.

Spine and scalp recordings as a function of intensity. A model for changes during spinal cord monitoring.

Spinal cord monitoring has used both spine and scalp recordings as indicators of spinal cord integrity. The relative merits of spine or scalp recordings to predict the quality of the afferent volley in the somatosensory pathway were addressed in this study by using various stimulus intensities as a way to model alterations of the size of the afferent volley. The results were analyzed to determine the correlation of central recordings taken at the spine or scalp with peripheral recordings. Spinal cord monitoring with somatosensory evoked potentials has been achieved with recordings of signals generated by either the spinal cord or the somatosensory cortex. Spine recordings are thought to be more stable, yet little evidence exists to document this statement. Seven patients were studied in the course of standard intraoperative spinal cord monitoring. Responses were recorded at the popliteal fossa, thoracic epidural, cervical spine, and scalp to tibial nerve stimulation at intensities varying from 0.5 to 2.0 times muscle twitch threshold. Normalized amplitudes of the response at the popliteal fossa were used to reflect the magnitude of the afferent volley. The amplitudes of the popliteal fossa response showed a high correlation (r = 0.90) with normalized amplitudes of epidural and cervical spine responses and moderate correlation (r = 0.49) with normalized amplitudes of scalp responses. The width of the 95% confidence limits for the inverse prediction of the afferent volley from epidural and cervical responses was nearly a third narrower than that from scalp responses. At low stimulus intensities, scalp responses were consistently observed when spine responses were absent, and scalp responses had lower response thresholds than did spine responses. The latencies of the popliteal fossa responses were not well correlated with latencies of either the epidural or cervical responses. These correlation and inverse prediction data suggest that the size of an afferent volley may be predicted more accurately by spine responses than by scalp responses. The presence of scalp responses at intensities too low to elicit detectable spinal-level responses suggests that scalp responses may be considered a sensitive indicator of a minimal afferent volley.

Electrophysiological improvement after co-implantation of carbon filaments and fetal tissue in the contused rat spinal cord

The electrophysiological integrity of the adult rat spinal cord was assessed at the lumbar, lower cervical and cortical levels after the animals sustained a severe contusion injury at the mid-thoracic level (T8) and received either carbon filament cultured with fetal spinal cord tissue implants, fetal tissue implants, or carbon filament implants alone. Somatosensory evoked potentials (SSEPs) and motor evoked potentials (MEPs) were recorded from all animal groups at the end of the 8-week survival period. The results of this study demonstrate that the spinal cord injured animals that received carbon filament cultured with fetal spinal cord tissue implants had the highest degree of electrophysiological recovery, indicating that this combination plays an important role in promoting recovery after injury.

Anterior Neck Recording of Intraoperative Somatosensory-Evoked Potentials in Children

The authors have developed technique of using anterior neck derivations to record posterior tibial nerve N28 during operative somatosensory-evoked potential monitoring. This prospective study of 10 patients compared the ease of application of electrodes and stability of waveforms with "traditional" posterior neck recordings. Somatosensory-evoked potential monitoring has been used in children since the 1980s. A number of important factors, patient related and technical, are unique to this age group, which can cause difficulty with execution and interpretation. A major patient-related problem is unreliability of the cortical response in somatosensory-evoked potential monitoring because of the effect of inhalation anesthesia. This has been described as occurring more often in the pediatric group, perhaps because of the continuing maturation of the complex somatosensory system. Thus, the authors have relied heavily on the cervical potential to monitor spinal cord integrity. Recording the cervical response using the traditional "posterior" montage may be technically impossible if the electrodes lie within the operative field. Posterior tibial somatosensory-evoked potentials were performed according to the guidelines of the American Electroencephalographic Society. Additional recordings were obtained from two anterior neck sites. Multiple reference electrodes were used to evaluate which montage rendered the most reliable waveform. Optimal montage for recording an anterior neck potential was cricoid cartilage-Cz'. There were no significant differences in the onset latency or peak amplitudes between the anterior and posterior cervical recordings. The anterior neck derivation was found to be an acceptable adjunct to posterior montages, being stable, of comparable latency, and of stable amplitude. Electrodes are easier to apply, less invasive than nasopharyngeal or esophageal electrodes. This technique allows one to examine patients who previously could not be monitored.

A new technique using somatosensory evoked potential guidance during descending and thoracoabdominal aortic repairs.

Recent work in our laboratory has demonstrated the effectiveness of somatosensory evoked potentials (SEPs) in identifying the critical intercostal arteries (CICAs) for preserving spinal cord integrity during simulated aortic aneurysm repairs in the pig. Further studies have also demonstrated increased preservation of neurological function during prolonged aortic clamping if CICAs are perfused until ligation or clipping, as opposed to transaortic identification of back-bleeding intercostals and their subsequent ligation. We have developed a technique of repair of descending thoracic and thoracoabdominal aortic aneurysms and dissections that uses these principles. Since January 1993, 26 patients have undergone repair of their aortas using this new technique and SEP directed intercostal artery ablation. There were 22 (85%) long-term survivors among 10 thoracoabdominal and 16 descending aortic repairs. All patients with uncorrected abnormal SEP recordings developed paralysis; one patient who required reimplantation of an intercostal artery island into the aortic graft had normal neurological function postoperatively. Paraplegia was seen in only one of the surviving patients, but this patient had normal intra- and postoperative SEPs (4% false negative). Our experience suggests that SEP-guided obliteration of intercostal arteries while maintaining perfusion may be a useful approach to the surgical repair of descending and thoracoabdominal aortic disease.

Spinal antinociception mediated by a cocaine-sensitive dopaminergic supraspinal mechanism

The role of dopaminergic descending supraspinal processes in mediating the antinociceptive action of cocaine was studied in the rat using a combination of extracellular neuronal recording and behavioral techniques. Neurons in the superficial laminae (I-II) of the spinal dorsal horn with receptive fields on the tail were recorded in anesthetized rats using insulated metal microelectrodes. Stimulation of the receptive field with either high intensity transcutaneous electrical pulses or with an infrared CO 2 laser beam produced a biphasic increase in dorsal horn unit discharge. Conduction velocity estimates indicated that the early discharge corresponded to activity in Aδ whereas the late response corresponded to activity in C afferent fibers. Cumulative doses of cocaine (0.1–3.1 mg/kg i.v.) inhibited the late response to either electrical or laser stimulation in a dose-related manner. The early response to laser, but not electrical, stimulation was also suppressed by cocaine. Neurons in the spinal dorsal horn with receptive fields on the ipsilateral hindpaw were activated by natural noxious (pinch) or innocuous (tap) somatic stimulation. Cocaine selectively suppressed nociceptively evoked dorsal horn unit discharge. This antinociceptive effect was dose-related (0.3–3.1 mg/kg, i.v.) and antagonized by eticlopride (0.05–0.1 mg/kg, i.v.), a selective D2 dopamine receptor blocker. The same doses of cocaine failed to inhibit the responses of dorsal horn neurons to low threshold innocuous stimulation. Complete thoracic spinal cord transection eliminated the antinociceptive effect of cocaine on dorsal horn neurons and also eliminated the cocaine-induced attenuation of the tail-flick reflex. These data demonstrte that cocaine selectively inhibits nociceptive spinal reflexes and the nociceptive responses of dorsal horn neurons primarily by means of a D2 dopaminergic receptor mechanism. This antinociceptive effect of cocaine is independent of its local anesthetic activity and requires the integrity of the thoracic spinal cord, suggesting that the drug potentiates or activates supraspinal dopaminergic projections to the dorsal horn.

Intravenous Anesthetic Alterations on the Spinal-Sciatic Evoked Response in Swine

Sciatic evoked responses (ScER) are used to monitor the integrity of the anterior spinal cord during spinal surgery. To evaluate the effects of intravenous anesthetics on the lumbar spinal evoked potential (SpEP) and ScER, 28 swine were anesthetized with halothane and an intravenous infusion of ketamine. After experimental preparation, anesthesia was maintained with ketamine, and the swine were divided into four equal groups (n = 7). The evoked responses were generated by delivering an electrical stimuli to the interspinous ligaments caudal to T1 and T2. The lumbar SpEP was recorded from the interspinous ligaments caudal to L2 and L3. The ScERs were recorded bilaterally from electrodes placed adjacent to the sciatic nerves. In Group 1 the effects of bolus dose administration of ketamine were evaluated. In Groups 2-4 anesthesia was maintained with ketamine by continuous infusion and the effects of fentanyl, sufentanil, and propofol on the L2 SpEP and ScER were evaluated. There was no significant change in either the amplitude or latency of the L2 SpEP with any anesthetic studied. Ketamine administered in bolus doses (2.5, 5, and 10 mg/kg) did not alter the ScER. After bolus doses of fentanyl 10 micrograms/kg or sufentanil 1 micrograms/kg, there was a similar decrement in the amplitude of the ScER (P < 0.05) that recovered at 10 min without changes in latency. There were no changes in the ScER associated with infusion of fentanyl 1-5 micrograms.kg-1 x h-1 or sufentanil 0.1-0.5 microgram.kg-1 x h-1.(ABSTRACT TRUNCATED AT 250 WORDS)

Physiologic and anesthetic alterations on spinal-sciatic evoked responses in swine.

Spinal-sciatic evoked responses (ScER) have been used successfully to monitor the integrity of the anterior spinal cord during spinal surgery. To evaluate the effects of hypercarbia, hypocarbia, induced hypotension, and hypothermia on the ScER, ten swine anesthetized with ketamine were subjected to varying levels of PaCO2, hypothermia, and induced hypotension. During variation of one physiologic variable, the other variables were closely regulated. There were no significant changes associated with variations in PaCO2. Decreasing temperature provided a consistent increase in latency (r = -0.78, P < 0.001) with no significant alteration in amplitude. Graded hypotension caused little increase in latency (3.2% at 30 mm Hg). The amplitude decrease averaged 23% at 60 mm Hg with a maximal decrease of 50% at 30 mm Hg. To study inhaled anesthetics, 21 swine anesthetized with ketamine were subjected to nitrous oxide (50% and 70%). After termination of the nitrous oxide, one of the potent inhaled anesthetics (n = 7 each) was administered in 0.25 minimum alveolar concentration (MAC) increments. Nitrous oxide caused a significant decrease in amplitude (average 43% and 61% at 50% and 70%) with minimal changes in latency. There was a dose-dependent decrease in amplitude and increase in latency with all inhaled anesthetics. The ScER disappeared at 1.0 MAC with all anesthetics. There were no differences between effects of equipotent concentrations of inhaled anesthetics. These findings may be helpful in the interpretation of the ScER response during anesthesia and surgery.

Sensory Evoked Potentials for Selective Monitoring of the Rat Spinal Cord: A Cerebellar Evoked Potential to Assess Ventral Cord Integrity

The purpose of this study was to evaluate two types of ascending sensory evoked potentials (SEPs) in the rat and their capacity for selective monitoring of dorsal versus ventral spinal cord integrity. SEPs were elicited by direct sciatic nerve stimulation. A cerebellar evoked response was recorded over the paramedian lobule of the cerebellar hemisphere (CEPpml) while somatosensory evoked potentials (SSEPs) were simultaneously recorded over the sensorimotor cortex. All components of the CEPpml and SSEP except the longest latency positive waves were present in each animal. At stimulus intensities of 3 to 10 mA, no significant changes in latency of the peaks were observed, but amplitudes of the longer latency responses tended to increase throughout this stimulation range. Unilateral sciatic stimulation resulted in bilateral cortical responses, larger ipsilaterally for the CEPpml, and contralaterally for the SSEP. Selective spinal cord lesions demonstrated N9 and P14 of the CEPpml to be mediated primarily through the ventral spinal cord, while P14 and N19 of the SSEP were conducted primarily through the dorsal columns. Sectioning of the cerebellar peduncles abolished N9 and P14 of the CEPpml despite persistence of the SSEP. This study demonstrates that selective assessment of the ventral and dorsal spinal cord is possible in the rat by monitoring SEPs.

Magnetic motor evoked potentials for assessing spinal cord integrity in dogs with intervertebral disc disease.

Magnetically elicited transcranial motor evoked potentials (MEPs) were studied in 37 dogs with type 1 intervertebral disc (IVD) disease. The waveforms were recorded from both cranial tibial muscles before and after surgery. The latencies and amplitudes obtained were compared to those of a control population (n = 14). MEPs were recordable in all dogs with mild or no neurologic deficits, but they were recordable in only 50% of ambulatory dogs that were severely ataxic. MEPs could not be elicited from nonambulatory dogs. There was a significant attenuation of the amplitudes in all clinical cases, even if the dogs demonstrated back pain alone. Significantly prolonged latencies were associated with neurologic deficits. MEPs were not good predictors of neurologic recovery. The responses obtained from the side where the disc material was found were not different from those recorded from the opposite side. MEPs were very sensitive to lesions of the spinal cord, as indicated by the significant changes in the waves in patients with mild or no neurologic deficits and in the loss of response in dogs that still demonstrated purposeful movement. Neurologic exams provided more accurate diagnoses and prognoses than did MEPs in dogs with IVD disease.

A multimodality approach lengthens warm ischemic time during aortic cross-clamping

To evaluate the effects of exsanguination, cerebrospinal fluid drainage (CSFD), steroids alone and in conjunction with CSFD on spinal cord perfusion pressure (SCPP), and neurological outcome following 70 min of normothermic spinal cord ischemia, we monitored proximal (Px BP) and distal (Ds BP) aortic blood pressure, cerebrospinal fluid pressure, and somatosensory evoked potentials (SEP) in 29 mongrel dogs. In all animals Px BP during aortic cross-clamping was controlled with partial exsanguination (40–50% circulating blood volume). Dogs were randomized into four groups (gp): gp 1 ( n = 6) control; gp 2 ( n = 8) steroids only (methylprednisolone 30 mg/Kg 10 min before aortic occlusion and 4 hr later); gp 3 ( n = 8) CSFD only; gp 4 ( n = 7) steroids and CSFD. Partial exsanguination effectively controlled Px BP during aortic crossclamping in all groups. After the statistically significant decrease from preclamp values, mean Px BP did not differ among groups (78.9, 81.2, 80.5, and 80.3 mm Hg, respectively, P > 0.05). Mean Ds BP decreased from systemic values to 12.6, 16.8, 16.7, and 17 mm Hg, respectively, after aortic occlusion ( P < 0.05); these values did not differ from one another. CSFP did not change significantly from its baseline value while the aorta was cross-clamped in gp 1; CSFP was significantly reduced to 6.2 mm Hg in gp 2, steroid-treated animals ( P < 0.05 vs gp 1); a further significant reduction in CSFP was noted in gp 3 and 4 undergoing CSFD (0.07 and 0.67 mm Hg, respectively, P < 0.05 vs gp 1 and 2). Spinal cord perfusion pressure was 1.8 mm Hg during aortic cross-clamping in gp 1 and remained significantly lower than gp 2, 3, and 4 (10.6, 16.4, and 16.5 mm Hg, respectively) throughout the experimental interval. Time until SEP loss was significantly longer in gp 2, 3, and 4 (14:40, 16:36, 17:48 min:sec, respectively) compared to gp 1 (7 min and 25 sec, P < 0.05). Somatosensory evoked potentials failed to return in all gp 1 animals, whereas early (within 30 min of reperfusion) postoperative return was noted in all animals in gp 2, 3, and 4 ( P < 0.05 vs gp 1). The incidence of paraplegia was significantly higher in gp 1 compared to gp 3 and 4 ( 5 6 vs 0 5 , and 0 6 , P < 0.05). This study indicates that exsanguination alone cannot maintain SCPP required to preserve spinal cord integrity during seventy minutes of normothermic ischemia. The addition of CSFD and steroids lengthens the time required to lose SEP and significantly reduces the risk of postoperative paraplegia, whereas steroids alone are not uniformly protective.

Relationship Among Level of Distraction, Evoked Potentials, Spinal Cord Ischemia and Integrity, and Clinical Status in Animals

Somatosensory evoked potentials (SEPs) and neurogenic-motor evoked potentials (NMEPs) were recorded after overdistraction of the spinal cord at T5-T6, T12-L1, or L3-L4. Measures of spinal cord perfusion and clinical status were also administered. Results indicated that stiffer spinal segments allowed less distraction than more flexible segments. SEPs and NMEPs were lost quickly after overdistraction in stiff segments and slowly in more flexible segments. However, SEPs were less sensitive than NMEPs to effects from overdistraction. Spinal cord perfusion and integrity were consistent with reduced perfusion and structural damage after overdistraction in stiff segments; extremely reduced perfusion but no structural changes in more flexible segments. The application of these results to the clinical situation was provided.

Relationship Between Duration of Spinal Cord Ischemia and Postoperative Neurologic Deficits in Animals

Stagnara wake-up tests, blood flow measures, somatosensory evoked potentials (SEPs), and neurogenic-motor evoked potentials (NMEPs) were elicited from 20 hogs before and after spinal cord overdistraction at L3-L4. Overdistraction was maintained from 5 to 30 minutes after loss of NMEPs. Results suggest that the longer the duration of overdistraction the greater the likelihood of paraplegia. Blood flow measures indicated that reduced perfusion was greatest at the distraction site but extended proximally and distally. Finally, NMEPs were more sensitive to onset of overdistraction and a more valid indicator of paraplegia than SEPs. NMEPs should provide the surgeon with more time for initiation of intervention techniques than SEPs. Because NMEPs and SEPs provide information regarding different spinal cord tracts, the authors continue to use both methods for monitoring the functional integrity of the human spinal cord during corrective spine surgery.

Effect of regional spinal cord blood flow and central control in recovery from spinal cord injury.

Forty-two cats were subjected to decerebration, thoracic and lumbar laminectomies, and isolation of the sciatic nerves. Spinal evoked potentials in response to bilateral sciatic nerve stimulation were recorded at L-3, and the spinal cord blood flow (SCBF) was measured by the hydrogen clearance technique. Thoracic cordotomy did not alter the lumbar SCBF or the central conduction time as determined by spinal evoked potentials. Thoracic cordotomy significantly lowered the lumbar spinal cord injury threshold. Continuous sciatic nerve stimulation increased the lumbar SCBF in normal and traumatized animals; however, it did not affect the spinal cord injury threshold as measured by recovery of the spinal evoked potentials. It appears that rostral spinal cord integrity is far more significant in recovery from spinal cord injury than the maintenance of regional SCBF.

Graded Spinal Cord Injuries Produced in Rabbits with Non-Invasive Microwave Hyperthermia

The use of non-invasive microwave energy to produce spinal cord injuries with intraspinal hyperthermia was studied in experimental animals. Lesions were produced with external beam microwave irradiation at 915 MHz in rabbits, using intraspinal temperature levels from 40 to 43 degrees C., and periods of heating ranging from 15 to 60 minutes. The parameters which determined thermal dose were the degree of temperature elevation in the spinal cord relative to the body core and the duration of that elevation. Thermal dose-response relationships were established by monitoring intraspinal temperatures during heating using an epidural thermistor probe at T8. Animals were examined 48 hours after lesion production and assigned a neurological grade. Injuries were grouped clinically according to their degree of relative functional severity as minimal, mild, moderate, or severe. Evaluation of spinal cord integrity was carried out by recording cortical somatosensory evoked responses (SER) following sciatic nerve stimulation. Increased SER latencies were first observed after heating the spinal cord to 41 degrees for 60 minutes. Impulse transmission was absent after heating to 42 degrees for 30 minutes, a thermal dose which produced complete paraplegia. Morphologically, lesion size and configuration were directly related to the thermal dose used in their production. Low thermal doses produced white matter edema limited to the posterior columns, while larger doses resulted in demyelination, retrograde neuronal changes, and infarction of the dorsal half of the cord. High thermal doses also produced foci of hemorrhage in the gray and white matter of the dorsal cord. These studies suggest that reproducible spinal cord injuries with predictable levels of neurological severity can be produced by noninvasive microwave heating.(ABSTRACT TRUNCATED AT 250 WORDS)

Cortical Evoked Potential Monitoring During Spinal Surgery

Cortical somatosensory evoked potentials (EP) were used for intraoperative monitoring of spinal cord integrity during 158 consecutive cases involving surgical correction of spinal deformities. Using a monitoring technique designed to optimize recording, 88% of patients had stable EP results at the time of spinal manipulation. However, five of six patients with preexistent neurological deficits did not have stable EPs. A persistent EP amplitude attenuation of greater than or equal to 40% was a sensitive and specific criterion for alarm. Seven patients had transient attenuations of greater than or equal to 40% after distraction that recovered, so that the distraction instrumentation could be left in place. Overall, this particular cortical somatosensory EP recording technique appears to be appropriate for clinical monitoring. Other recording techniques may be necessary for patients with preexisting neurological deficits.