| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Intraoperative recording of myogenic motor responses evoked by transcranial electrical stimulation (tcMEPs) is a method of assessing the integrity of the motor pathways during aortic surgery. To identify conditions for optimal spinal cord monitoring, we investigated the effects of manipulating the level of neuromuscular blockade (T1 response of the train-of-four (TOF) stimulation 5%–15% versus T1 response 45%–55% of baseline), as well as the number of transcranial pulses (two versus six stimuli) on the within-patient variability and amplitude of tcMEPs. Ten patients (30–76 yr) scheduled to undergo surgery on the thoracic and thoracoabdominal aorta were studied. After achieving a stable anesthetic state and before surgery, 10 tcMEPs were recorded from the right extensor digitorum communis muscle and the right tibialis anterior muscle in response to two-pulse and six-pulse transcranial electrical stimulation with an interstimulus interval of 2 ms during two levels of neuromuscular blockade. The right thenar eminence was used for recording the level of relaxation. The tcMEP amplitude using the six-pulse paradigm was larger (P < 0.01; leg and arm) compared with the amplitude evoked by two-pulse stimulation during both levels of relaxation. The within-patient variability, expressed as median coefficient of variation, was less when six-pulse stimulation was used. At a T1 response of 45%–55% of baseline, larger, less variable tcMEPs were recorded than at a T1 response of 5%–15%. Our results suggest that the best quality of tcMEP signals (tibialis anterior muscle) is obtained when the six-pulse paradigm is used with a stable level of muscle relaxation (the first twitch of the TOF—thenar eminence—at 45%–55% of baseline).
Implications: This study shows that six-pulse (rather than two-pulse) transcranial electrical stimulation during a stable anesthetic state and a stable neuromuscular blockade aimed at 45%–55% (rather than 5%–15%) of baseline provides reliable and recordable muscle responses sufficiently robust for spinal cord monitoring in aortic surgery.
Spinal cord integrity can be monitored by the intraoperative recording of transcranial myogenic motor evoked responses (tcMEPs). Volatile anesthetics (1–5), propofol (6,7), benzodiazapines (6–8), and nitrous oxide (9–11) are potent depressants of the tcMEP. Neuromuscular blocking drugs also decrease the tcMEP response (12). Most anesthetic regimens may therefore lead to a misinterpretation of the tcMEP. Thoracoabdominal aortic aneurysm surgery consequently demands a stable level of neuromuscular blockade to minimize the occurrence of false-positive changes during surgery and a neuromuscular blockade sufficiently deep to avoid suboptimal operating conditions. The effects of different levels of neuromuscular blockade on the variability of MEP monitoring are unknown. Multiple transcranial electrical stimulation may overcome the neuromuscular drug effect and may reduce the within-patient variability of the tcMEP. First, the aim of this study was to quantify the within-patient variability and amplitude of the tcMEPs in response to two-pulse and six-pulse transcranial electrical stimulation during stable fentanyl/nitrous oxide anesthesia with a small-dose propofol infusion. Second, we studied the variability at two levels of partial neuromuscular blockade to identify conditions that reduce the within-patient variability in patients at risk.
Ten patients (2 female, 8 male, aged 30–76 yr) undergoing surgery for thoracic or thoracoabdominal aortic aneurysm gave their written, informed consent to participate in this study, approved by our institutional ethics committee. All patients were free from any neuromuscular disorder and epilepsy. The patients were premedicated with morphine 10 mg and haloperidol 5 mg IM 1 h before surgery. After venous cannulation of the right arm and the right radial artery, anesthesia was induced with diazepam 0.2–0.3 mg/kg and fentanyl 20–30 µg/kg, and a small-dose propofol infusion was started to maintain a plasma steady-state concentration of (target-controlled infusion) 1 µg/mL; infusion rate 150–200 mg/h or 40–50 µg · kg–1 · min–1 after a bolus of approximately 20–40 mg in 3 s. Tracheal intubation was performed using a left-sided double-lumen tube positioned with the aid of a fiberoptic bronchoscope. If necessary, succinylcholine 20 mg was given IV to facilitate intubation. Controlled ventilation was adjusted to maintain normocapnia (ETCO2 4.0–4.5 kPa) and to administer nitrous oxide 50% in oxygen. A pulmonary artery catheter was inserted via the right internal jugular vein. A nasogastric tube, an indwelling bladder catheter, a rectal thermometer, and a muscle thermometer (right tibialis anterior muscle) were placed. The right-sided thenar eminence was used to monitor the neuromuscular blockade. Before administration of a neuromuscular drug, the compound muscle action potential baseline was obtained from the thenar eminence after supramaximal stimulation of the median nerve at the wrist using a general evoked response stimulator triggered from a personal computer. A mivacurium chloride infusion was used to maintain the first twitch of the train-of-four (TOF) at 5%–15% or 45%–55% of baseline. The T1 response was displayed on the computer screen. The patient was positioned on a beanbag in the right lateral decubitus position, and two intrathecal catheters (one for monitoring intrathecal pressure, one for drainage of cerebrospinal fluid) were placed via the second and third lumbar interspaces. Spinal fluid drainage was continued throughout the procedure to maintain the pressure below 10 mm Hg. Routine anesthetic monitoring for major vascular surgery was performed and electronically recorded every 30 s. We created three phases with different levels of neuromuscular blockade (NMB). At first, the T1 response was maintained at 100% of baseline. Subsequently, the T1 response was maintained at 5%–15% of baseline or at 45%–55% of baseline in order of occurrence. In every phase, anodal two-pulse and six-pulse electrical stimulation was delivered at the vertex with the cathode at the Fpz position (International 10–20 system for the placement of the electroence phalogram electrodes). Before surgery, custom-made disc electrodes of a silver-copper alloy with a diameter of 25 mm filled with conduction paste (carmellose sodium 200–800 mPa.s 1.35 g, glycerol 100% 8.50 g, sodium chloride 7.00 g, potassium acid tartrate 0.375 g, and distilled water ad 50.0 g) were attached to the scalp. To improve electrode stability, gauze pads were fixed over the discs with collodium. Both disc electrodes were connected to a multipulse electrical stimulator. The trains of two- and six-pulse stimuli with an interstimulus interval of 2 ms were triggered from a computer every minute. The output voltage was set at 490 V, guaranteeing descending volleys down the cord of maximal amplitude. The stimulation device used has an internal safety circuit that does not allow higher output voltages. The pulse duration was 50 µs. The study was performed before any surgical action was undertaken to avoid interference that might have resulted in impaired spinal cord functioning. After this initial period of measurement recording, which lasted approximately 60 min, surgery proceeded as usual for total thoracoabdominal aortic replacement. The myogenic potentials were recorded on a timebase of 150 ms, traversing a bandpass filter of 50 Hz–500 Hz and amplified 10,000 times. The signals were sampled at 1 kHz, digitized with a 12-bit A/D converter, and displayed on a standard computer screen. The amplitude was measured as the voltage from the most negative component to the most positive component of the evoked electromyographic activity (tcMEP). The area under the curve (AUC) of the tcMEP was also calculated. tcMEPs were recorded from the upper and lower extremities at the right side of the body, i.e., the extensor digitorum communis muscle and the tibialis anterior muscle, because, typically, only the right lower limb is perfused continuously using the left heart bypass. Standard silver-silver chloride disc electrodes (1 cm in diameter; Pomed Implements, Apeldoorn, The Netherlands) filled with conducting paste were placed on the muscle belly and on the tendon of those two muscles. In every phase, 10 tcMEPs after two pulses and 10 tcMEPs after six pulses were collected. Every phase lasted 15–20 min. For each phase, the mean amplitudes and mean AUC were calculated from 10 consecutive responses. The median value of 10 means of the 10 patients were also derived. To estimate the within-patient variability, we determined the coefficient of variation (CV) (SD/mean*100%) of 10 consecutive responses in each phase. The median value of 10 CVs of the 10 patients was also noted. Intraindividually, there was a normal distribution in the measured tcMEP variables. Interindividually, there was a skewed distribution in the tcMEP amplitudes and tcMEP AUCs. Differences in amplitude and AUC and the CVs (to analyze the within-patient variability) were compared between the two-pulse and six-pulse paradigm, as well as among the different levels of neuromuscular blockade, using analysis of variance (ANOVA) for repeated measurements in SPSS (Statistical Package for the Social Sciences, Inc, Chicago, IL). This ANOVA provides point estimates with their 95% confidence intervals. Knowing their non-Gaussian distribution, the amplitudes, AUCs, and their CVs in each phase were also compared using the Friedman statistic. A P value <0.05 was considered significant. Each patient was neurologically examined and questioned, if possible, for recall of intraoperative events the first postoperative day in the intensive care unit. If the patient was properly sedated, tcMEPs were recorded the first postoperative day to assess the integrity of the motor pathways.
No neurological sequellae occurred after trains of two- and six-pulse transcranial electrical stimulation. No burn marks were observed at the stimulation sites. During the recording period, all patients were hemodynamically stable, the systolic blood pressure was maintained between 80 and 100 mm Hg, and the pulse rate was maintained between 50 and 100 bpm. There was no significant change in measured temperatures (36.3 ± 0.5°C) and no significant change in ETCO2 (4.0 ± 0.3 kPa) during the course of the study. There was no suspicion of awareness, and no patients recalled intraoperative events. Figure 1 shows box plots of tibialis anterior muscle tcMEP amplitude and extensor digitorum communis muscle tcMEP amplitude after two-pulse and six-pulse electrical stimulation during no neuromuscular blockade, T1 at 45%–55% of baseline and T1 at 5%–15% of baseline. Box plots of the CV (SD/mean*100%) of the tcMEP amplitude are also shown.
Table 1 shows the measured values of the tcMEP amplitude and AUC of the right tibialis anterior muscle. Table 2 shows the data of the right extensor digitorum communis muscle. Large interpatient variability was observed.
Table 3 shows the mean differences of the measured variables (and P values in parentheses) comparing the different levels of neuromuscular blockade and the pulse paradigms. The results of the Friedman statistics analyzing the amplitude and AUC are displayed in Tables 1 and 2, as well as in Figure 1 for the amplitude.
Compared with the two-pulse paradigm, the six-pulse paradigm resulted in statistically significant larger responses in amplitude (P < 0.01) and in AUC (P < 0.01) obtained from the leg and arm across all levels of neuromuscular blockade. There was a significant decrease in the CV for the tibialis anterior muscle tcMEP AUC (P < 0.05) between the two paradigms. The least variable amplitudes (P = 0.05) were observed in the leg during six-pulse stimulation. Comparing the pulse stimulation paradigms, no difference was found in variability of the measured tcMEP variables obtained from the arm.
No significant difference was found in amplitude and AUC of the tcMEP in the tibialis anterior muscle among the levels of NMB. The recordings in the extensor digitorum communis muscle demonstrated a reduction in amplitude and AUC (P In the leg, the least variable amplitudes and AUCs were observed during a NMB with the T1 of TOF aimed at 45%–55% after six-pulse stimulation. The CV (median value) of the tcMEP amplitude in the tibialis anterior muscle was 11%. Compared with a T1 of 5%–15%, there was a statistically significant smaller CV (P < 0.05) in the leg during a NMB aimed at a T1 of 45%–55%. In the arm, the least variable amplitudes were recorded during a NMB T1 aimed at 45%–55% after six-pulse stimulation. The CV (median value) of the tcMEP amplitude in the extensor digitorum communis muscle was 31%. During the course of surgery, the evoked responses were maintained, and all patients awoke in the intensive care unit without neurologic deficits.
Monitoring the motor pathways of the spinal cord during operations for thoracic and thoracoabdominal aortic aneurysms provides a method for detecting ischemia of the anterior cord that may not affect somatosensory evoked potential recordings. To accomplish highly sensitive and highly specific spinal cord monitoring, it is important to have a constant anesthetic regimen and a constant level of NMB, which results in the least spontaneous amplitude variability. Variability of tcMEPs occurs because the fibers of the stimulated muscle are not consistently recruited by the electrical stimulus. This means that the number and timing of excited motor units participating in the myogenic response (i.e. the amplitude and AUC) differs among stimuli. Considering the motor pathways in a patient under anesthesia and during partial NMB, this changing number of excited motor units may be caused by fluctuations of motor neuron excitability and neuromuscular transmission. Because both propofol and nitrous oxide anesthesia interfered equally with the tcMEP variables throughout measurement recording, this study shows that the use of six-pulse stimulation, compared with two-pulse stimulation, results in a significantly augmented tcMEP amplitude and AUC in the arm and leg. The CV of the tcMEP was less after six-pulse stimulation. Temporal summation of excitatory inputs to the motor neuron pool, thereby increasing synaptic efficacy, probably results from six-pulse stimulation of the motor cortex, which provides less variable augmented responses and more reliable recordings than two-pulse transcranial electrical stimulation (13–15). Longer pulse trains will increase cortical stimulation and, consequently, have variable effects on the number and types of corticospinal tract cells activated due to cortical stimulus spread within the cortex. In this study, the tcMEP amplitude variability was less for the tibialis anterior muscle compared with the extensor digitorum communis muscle. This difference may be due to a higher threshold of activation of the extensor digitorum communis muscle, because the placement of the stimulating electrodes was optimal for tcMEP recording from the leg muscles, not from the arm muscles (16). Another possibility is that multipulse cortical stimulation results in further different interneuronal synaptic and direct monosynaptic input to the anterior horn cells supplying the muscles of the arm and leg from which myogenic recordings took place. The degree of NMB and two- versus six-pulse stimulation may lead to variability in mobilization or depletion of acetylcholine at the motor endplate. Our results show that the tcMEP response decreases and its variability (expressed as SD/mean*100%) increases with increasing levels of NMB. Muscle relaxants block the transmission from the peripheral nerve to the muscle fibers. The tcMEP amplitude reduction is caused by a reduction in the number of motor units responding to the stimulus. There is a major difference in sensitivity to NMB drugs among muscle groups (17), as shown in this study among the tibialis anterior muscle, the extensor digitorum communis muscle, and the thenar eminence. NMB sensitivity is not related to blood flow (17), but to the unique physical and architectural features, as well as intrinsic physiological and biochemical properties, of each muscle (18). Furthermore, a change in NMB may affect the afferent input (Ia activity) from the peripheral nervous system to the dorsal horn. Properly timed somatosensory stimulation modulates tcMEP, facilitating production of higher amplitudes (19–21). From this study, we cannot state whether the observed changes in motor neuronal excitability is related to different afferent input from muscle spindles as a consequence of different levels of NMB. Apart from the physiologic possibilities, the increase in variability seen with increasing NMB may simply reflect an increasingly unfavorable signal to noise ratio. Our data are clinically relevant because it is important to know the within-patient variability of the arm tcMEP. These data can be used to differentiate among the various causes of lower extremity tcMEP loss. Upper and lower tcMEPs were recorded simultaneously to allow differentiation among the various causes of lower extremity tcMEP loss. Further studies are required to optimize the efficiency of stimulus delivery, resulting in sufficient reproducible muscle responses in both the arm and leg. In conclusion, using the described anesthetic regimen, the tcMEP amplitude evoked by paired stimulation is more variable than after trains of six-pulse stimulation. The application of a stable level of NMB with a T1 response aimed at 45%–55% (thenar eminence) of baseline seems to provide the best quality of intraoperative tcMEPs during fentanyl/nitrous oxide anesthesia and small-dose propofol infusion.
The authors thank C. J. Kalkman for assistance in the preparation of this manuscript and S. J. Hengeveld for technical assistance.
Presented in part at the Seventh International Symposium on Spinal Cord Monitoring, March 18–20, 1998, Osaka, Japan.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|