J Clin Monit Comput DOI 10.1007/s10877-014-9571-9
ORIGINAL PAPER
Intraoperative neurophysiological monitoring during spine surgery with total intravenous anesthesia or balanced anesthesia with 3 % desflurane Tod B. Sloan • J. Richard Toleikis Sandra C. Toleikis • Antoun Koht
•
Received: 14 September 2013 / Accepted: 12 March 2014 Ó Springer Science+Business Media New York 2014
Abstract Total intravenous anesthesia (TIVA) with propofol and opioids is frequently utilized for spinal surgery when somatosensory evoked potentials (SSEPs) and transcranial motor evoked potentials (tcMEPs) are monitored. Many anesthesiologists would prefer to utilize low dose halogenated anesthetics (e.g. 1/2 MAC). We examined our recent experience using 3 % desflurane or TIVA during spine surgery to determine the impact on propofol usage and on the evoked potential responses. After institutional review board approval we conducted a retrospective review of a 6 month period for adult spine patients who were monitored with SSEPs and tcMEPs. Cases were included for the study if anesthesia was conducted with propofol–opioid TIVA or 3 % desflurane supplemented with propofol or opioid infusions as needed. We evaluated the propofol infusion rate, cortical amplitudes of the SSEPs (median nerve, posterior tibial nerve), amplitudes and stimulation voltage for eliciting the tcMEPs (adductor pollicis brevis, tibialis anterior) and the amplitude variability of the SSEP and tcMEP responses
T. B. Sloan (&) Department of Anesthesiology, Anschutz Office West (AO1), MS 8202, University of Colorado Denver School of Medicine, 12631 E 17th Avenue, Aurora, CO 80045, USA e-mail: [email protected] J. R. Toleikis Department of Anesthesiology, Rush Medical College, Rush University Medical Center, Chicago, IL, USA S. C. Toleikis Department of Anesthesiology, Rush University Medical Center, Chicago, IL, USA A. Koht Departments of Anesthesiology, Neurosurgery, and Neurology, Northwestern University, Chicago, IL, USA
as assessed by the average percentage trial to trial change. Of the 156 spine cases included in the study, 95 had TIVA with propofol–opioid (TIVA) and 61 had 3 % expired desflurane (INHAL). Three INHAL cases were excluded because the desflurane was eliminated because of inadequate responses and 26 cases (16 TIVA and 10 INHAL) were excluded due to significant changes during monitoring. Propofol infusion rates in the INHAL group were reduced from the TIVA group (average 115–45 lg/kg/min) (p \ 0.00001) with 21 cases where propofol was not used. No statistically significant differences in cortical SSEP or tcMEP amplitudes, tcMEP stimulation voltages nor in the average trial to trial amplitude variability were seen. The data from these cases indicates that 1/2 MAC (3 %) desflurane can be used in conjunction with SSEP and tcMEP monitoring for some adult patients undergoing spine surgery. Further studies are needed to confirm the relative benefits versus negative effects of the use of desflurane and other halogenated agents for anesthesia during procedures on neurophysiological monitoring involving tcMEPs. Further studies are also needed to characterize which patients may or may not be candidates for supplementation such as those with neural dysfunction or who are opioid tolerant from chronic use. Keywords Propofol Desflurane Total intravenous anesthesia Somatosensory evoked potentials Motor evoked potentials Spinal surgery
1 Introduction Total intravenous anesthesia (TIVA) is often utilized when electrophysiological monitoring is used during spine surgery. This is frequently accomplished using infusions of propofol and an opioid (e.g. sufentanil). Anesthesiologists
123
J Clin Monit Comput
often would prefer to supplement this with low dose (e.g. 1/2 MAC) of a halogenated anesthetic agent (e.g. desflurane) in these cases because of their contribution to the goals of anesthesia and as well as a means to conserve propofol. The use of an inhalational agent such as desflurane with intravenous anesthetics may be particularly helpful in patients with opioid tolerance since the halogenated agents work through synaptic mechanisms different than the opioids [1]. Because both propofol and halogenated agents have been associated with amplitude reduction of cortical somatosensory evoked potentials (SSEPs) and muscle recordings from transcranially evoked motor evoked potentials (tcMEPs), the comparative impact on monitoring needs to be characterized. To our knowledge no systematic study of propofol–opioid anesthesia versus 1/2 MAC of desflurane with propofol and/or opioid infusions has been published. The potential impact of inhalational agents on tcMEPs has been specifically raised by Woodforth who indicated tcMEP responses may be ‘‘difficult to rely on’’ because of marked variations in their amplitude when nitrous oxide and isoflurane were used [2]. Hence the variability of the responses needs to be characterized when inhalational agents are added since this may change the usefulness of amplitude criteria during monitoring. We retrospectively reviewed our experience with spine surgery cases when propofol–opioid anesthesia and 1/2 MAC desflurane with propofol and/or opioid infusions were utilized in order to determine the impact of the halogenated agent on cortical SSEPs, multipulse tcMEP recordings, and the utilized dose of propofol.
2 Methods After Institutional Review Board approval, the intraoperative monitoring records for a 6 month period between September 1, 2012 and March 31, 2013 were reviewed to identify spinal surgeries when intraoperative monitoring was utilized. The anesthesia records of these cases were used to identify procedures when the anesthesia management utilized only propofol–opioid infusions (TIVA group) or 3 % desflurane with infusions of propofol and/or opioid (INHAL group). The infusion rates of propofol during the majority of the surgical period were recorded from the electronic medical record. We excluded cases where other intravenous agents (e.g. dexmeditomidine, ketamine, and lidocaine) or other inhalational agents (e.g. sevoflurane, nitrous oxide) were used. Other recorded data included the surgical procedure, case duration, patient gender, age, and weight. During the selected investigative period, several individuals provided anesthesia services. No particular assignment schedule was followed and the use of halogenated
123
agents was decided upon by the individual providers. When desflurane was used, it was routinely utilized during the procedure at 3 % expired (1/2 minimal alveolar concentration: MAC) with infusions of propofol and/or opioids as needed. Cases where desflurane was used briefly at induction and eliminated prior to monitoring were considered for inclusion in the TIVA group. Since it is known that inhalational agents (especially when accompanied with nitrous oxide) can markedly depress the amplitude or ability to acquire SSEP or tcMEP responses, those cases where we were unable to acquire responses in the presence of desflurane and the desflurane was terminated for monitoring were not included in the case series. The rates of propofol infusion were adjusted clinically at the discretion of the anesthesiologist as were the choices and rates of opioid infusion (sufentanil or remifentanil). In these cases electrophysiological monitoring was conducted using a Cadwell Cascade Elite (Cadwell Laboratories, Inc., Kennewick, WA). Cortical SSEPs were recorded with sterile subdermal needles placed at C30 , C40 , Cz0 and referenced to Fz using bandpass filtration of 30-500 Hz (all electrode positions based on the International 10–20 system). Each SSEP averaged response consisted of two hundred, 50 or 100 ms traces of cortical activity that resulted from stimulation of a median nerve (MN) at the wrist or posterior tibial nerve (PTN) at the ankle, respectively. Stimulation was provided through sterile subdermal needles at a frequency of 2.66 Hertz (Hz) using 100 ms pulses at a current intensity just above motor threshold. The cortical amplitudes of the SSEPs that resulted from median and posterior tibial nerve stimulation were recorded for each case. TcMEP responses were recorded from sterile subdermal needles placed in the abductor pollicis brevis and tibialis anterior muscles using bandpass filtration of 10–3,000 Hz. These responses were elicited by transcranial electrical stimulation at C1–C2 using a train of six pulses with a 50 ms pulse width and an interpulse interval of 2 ms. The stimulation voltage was chosen by the monitoring professional based on the intensity that was needed for response acquisition. For the purpose of analysis of the effect of the anesthesia on the SSEP amplitudes, the last monitored response prior to stopping or tapering the propofol and opioid infusions in anticipation of the conclusion of the procedure was used. When left and right-sided responses differed in amplitude, the response with the largest amplitude was chosen. Similar to the analysis of the SSEP amplitudes, the amplitudes of the last responses from the abductor pollicis brevis (APB) and tibialis anterior (TA) prior to terminal anesthetic change were recorded. Similar to that of the SSEPs, when the left and right-sided responses differed in amplitude, the response with the largest amplitude was selected. The
J Clin Monit Comput
stimulation voltages were also recorded at that time. If more than one stimulation intensity was used, the highest one was recorded. In order to quantitate the effect of the anesthesia on amplitude variability, the data were grouped into data sets for each patient limb. Hence, for each patient, separate SSEP data sets were created using all of the evoked response recordings for the left and right median and posterior tibial nerves (4 data sets per patient). Similarly, four additional data sets were created using the tcMEP amplitudes of the left and right adductor pollicis brevis and tibialis anterior muscles. For each data set, the average signal variability was calculated by determining the average of the absolute percentage change from each trial to the subsequent response. The median amplitude for each data set was also calculated to characterize the relationship of response variability and amplitude. Data sets where a monitoring alert occurred (e.g. a sustained 50 % or greater drop in the amplitude of any cortical SSEP response or a total loss of tcMEP responses) in conjunction with an operative event (e.g. surgical event, anesthetic bolus, etc.) were excluded from inclusion in the analysis. Chi square analysis was used to determine if there was a significant difference in the occurrence of alerts between the groups. The results were divided into two groups depending on the use of desflurane. The TIVA group utilized only propofol and opioid infusions (sufentanil or remifentanil) whereas for the INHAL group, desflurane was used throughout the case with infusions of propofol and/or opioid as determined by the anesthesiologist. The SSEP and tcMEP data are displayed using box plots because the values are not normally distributed. The box displays the middle half of the data (between the 25th and 75th percentile). The bar within the box represents the median of the data and the vertical line extends between the minimum and maximum values. A student’s t test was used to compare groups. p \ 0.05 was considered significant.
3 Results During the study period, 156 patients were identified for inclusion. Anesthesia was conducted using propofol–opioid infusions (TIVA, 95) or 3 % desflurane with propofol and/or
opioid infusions (INHAL, 61). Three cases were excluded from the INHAL group because desflurane was removed because of inadequate responses (see below) and 26 cases were excluded (TIVA 16, INHAL 10) because of monitoring alerts. The final case series included 127 cases (TIVA 79, INHAL 48) which included 15 anterior cervical procedures (TIVA 13, INHAL 2), 32 posterior cervical procedures (TIVA 23, INHAL 9), and 82 posterior thoracic and/or lumbar procedures (TIVA 43, INHAL 37). Table 1 shows the characteristics of the groups and the average propofol infusion rates. The difference between the propofol infusion rates was statistically significant (p \ 0.0001). Three patients in the INHAL group were excluded because the desflurane was terminated when the monitored responses were not discernable from the background noise. In one case there were no lower extremity SSEP responses in the presence of desflurane but the responses returned (average amplitude 0.3 lV) after termination of desflurane usage. In the second case, one lower extremity tcMEP was absent (the other limb response had an amplitude less than 50 lV). Bilateral responses were present (average amplitude 400 lV) when the use of desflurane was terminated. In the third case no lower limb tcMEP responses could be recorded in the presence of desflurane and the responses were still not discernable after desflurane was eliminated. Thirty-four intraoperative monitoring alerts occurred in 26 patients (TIVA 16, INHAL 10) which were excluded from analysis. Five alerts (TIVA 2, INHAL 3) occurred in five patients who had reversible amplitude decline or loss in the left leg response associated with occlusion of the left iliac artery during anterior exposure to the spine. Seven alerts (TIVA 4, INHAL 3) occurred in seven patients who had amplitude decline of one or both upper extremity SSEP responses during surgery on the thoracic or lumbar spine which was partially corrected by repositioning the arms. Six alerts (TIVA 3, INHAL 3) occurred in six patients where the SSEP amplitude decline occurred gradually over several hours eventually reaching 50 % or less. Fifteen alerts (TIVA 13, INHAL 2) were associated with surgical events in 12 patients where abrupt decline or loss of responses was associated with manipulation of a spinal cord tumor (5), hematoma removal (1), spinal cord traction (3), pedicle screw placement (2), dural tear (1), neck positioning (2), and corpectomy (1). One alert resulted from administration of neuromuscular blockade. The
Table 1 Basic case characteristics Gender (number M, F)
Age (years) (Average, min, max)
Weight (kg) (Average, SD)
Duration (min) (Average, SD)
Propofol (lg/kg/min) (Average, SD)
TIVA
41, 38
56 (22, 87)
79 (17)
345 (142)
115 (24)
INHAL
21, 27
59 (24, 83)
81 (24)
330 (123)
49 (49)
123
J Clin Monit Comput
occurrence of total alerts between groups was not statistically significant (p [ 0.94). Assessment of postoperative outcome revealed no new neurological deficits in any of the patients included in the data analysis. In addition, in the patients with alerts, no new neurological deficits were observed except in the four patients with the five alerts related to manipulation of spinal cord tumors (TIVA 3, INHAL 1) and hematoma removal (INHAL). The number of cases is too small to determine if the presence of the inhalational agent changed the sensitivity of the monitoring to the impending pathology.
Fig. 1 Median nerve and posterior tibial nerve cortical SSEP amplitudes. Box plot of SSEP cortical amplitudes from stimulation of the median nerve (MN, left) or posterior tibial nerve (PTN, right) where anesthesia was propofol–opioid TIVA (TIVA) or 3 % desflurane supplemented as needed with propofol and opioid infusions (INHAL). The maximum values are noted if they exceed the limits of the plot Fig. 2 Muscle amplitudes from transcranial motor evoked potential (tcMEP) stimulation. Box plot of tcMEP amplitudes from recordings from the adductor pollicis brevis (APB, left) or tibialis anterior (TA, right) where anesthesia was propofol–opioid TIVA (TIVA, left) or 3 % desflurane supplemented as needed with propofol and opioid infusions (INHAL, right). The maximum values are noted if they exceed the limits of the plot
123
Figure 1 shows the box plots of the amplitudes of the cortical SSEPs that resulted from median and posterior tibial nerve stimulation. The median values for the median nerve SSEP was 1.92 lV (TIVA) and 1.60 lV (INHAL) (p = 0.079). For the posterior tibial nerve the median values were 1.09 lV (TIVA) and 1.00 lV (INHAL) (p = 0.44). Figure 2 shows the box plots of the amplitudes of the tcMEP responses recorded from the APB and TA muscles. The median values for the APB was 2016 lV (TIVA) and 2,867 lV (INHAL) (p = 0.15) and for the TA was 812 lV (TIVA) and 749 lV (INHAL) (p = 0.82). Figure 3 shows the box plots of the stimulation intensities used for eliciting the tcMEP responses (median values 400 V for each group) (p = 0.10). No significant differences were detected between these groups. The results of the trial to trial variability are shown in Fig. 4. The median nerve and posterior tibial nerve cortical SSEP responses revealed an average trial to trial percentage change of 14–18 % with small but not statistically significant differences between groups (MN, p = 0.27) (PTN, p = 0.054). Average variability for individual limbs ranged from 2.9–42.3 % with larger values of variability occurring when the average amplitude of the responses was less than 4 lV and particularly when it was less than 2.5 lV (Fig. 5). Similarly, the amplitudes of tcMEP responses demonstrated an average 34–41 % inter-trial change in amplitude for which small but not statistically significant differences between groups (Fig. 6) were found (APB, p = 0.80) (TA, p = 0.66). Average variability for tcMEP responses from individual limbs ranged from 3 to over 300 % with larger variability in response amplitude occurring when the average amplitude was less than 4,500 lV and particularly when it was below 2,000 lV (Fig. 7).
J Clin Monit Comput
Fig. 5 Average cortical SSEP amplitude trial to trial percentage change versus average limb amplitude. Plot of average trial to trial percentage change of SSEP cortical amplitudes (MN and PTN combined) versus average amplitude where anesthesia was propofol– opioid TIVA (TIVA) or propofol–opioid-3 % desflurane (INHAL) Fig. 3 Transcranial motor evoked potential stimulation voltage. Box plot of tcMEP stimulation voltage used in the case series where anesthesia was propofol–opioid TIVA (TIVA, left) or 3 % desflurane supplemented as needed with propofol and opioid infusions (INHAL, right)
Fig. 4 Average trial to trial percentage change of cortical SSEP amplitude. Box plot of average trial to trial percentage change of SSEP cortical amplitudes from stimulation of the median nerve (MN, left) or posterior tibial nerve (PTN, right) where anesthesia was propofol–opioid TIVA (TIVA) or 3 % desflurane supplemented as needed with propofol and opioid infusions (INHAL). The maximum values are noted if they exceed the limits of the plot
4 Discussion These data demonstrate that 3 % desflurane can be used in some patients during the monitoring of cortical SSEPs and the muscle responses of tcMEPs. The cases in this series show that, as a group, the amplitudes of the cortical SSEPs, amplitudes of the tcMEP muscle responses, and the tcMEP
Fig. 6 Average trial to trial percentage change of motor evoked potential amplitude. Box plot of average trial to trial percentage change of tcMEP amplitudes from recordings from the adductor pollicis brevis (APB, left) or tibialis anterior (TA, right) where anesthesia was propofol–opioid TIVA (TIVA, left) or propofol– opioid-3 % desflurane (INHAL, right). The maximum values are noted if they exceed the limits of the plot
stimulation voltages are comparable. Further, as a group, the variability of the responses as measured by the average percentage change from trial to trial is also comparable. One explanation may be that the depression of the responses associated with 3 % desflurane is offset by the effect of the lower propofol infusions. Certainly this effect would be better characterized in a crossover study designed to compare the effects of the two agents. It should be noted that this observational study is likely underpowered to detect a significant difference between these groups but the
123
J Clin Monit Comput
Fig. 7 Average tcMEP amplitude trial to trial percentage change versus average limb amplitude. Plot of average trial to trial percentage change of tcMEP amplitudes (both APB and TA plotted) versus average amplitude where anesthesia was propofol–opioid TIVA (TIVA) or propofol–opioid-3 % desflurane (INHAL)
data suggests a lack of clinically significant difference in these patients. This is not the case in all patients since three patients had no recordable SSEP or tcMEP responses when 3 % desflurane was initially used and the anesthesia was switched to TIVA. With this switch the responses improved in two cases, particularly the tcMEP responses, suggesting that in their case the effect of desflurane exceeded the impact of a larger dose of propofol. Each of these patients had neurological symptoms (numbness, tingling in an extremity) which may have impacted on the relative anesthetic effect of desflurane and propofol. In one case responses could not be recorded without desflurane consistent with the observation that not all patients have recordable responses under TIVA despite clinical sensory and motor functional deficits. This underscores one limitation of this study since the patients where desflurane was used were not randomly chosen. Practitioners in this case series may have reserved desflurane usage for patients who did not have neural symptoms or a disease process which is associated with more difficult monitoring (numbness, tingling, weakness, diabetes, vascular disease), or where chronic opioid use was known to make anesthesia management more difficult. This suggests that there might be a clinically significant difference in the effect of the choice of anesthesia is some patients. In this case series 26 patients were excluded from analysis of the propofol infusion rate and the response variability because one or more monitoring alerts occurred during the case. Although the frequency of alerts were not different between the groups it is unknown if the type of
123
anesthesia changed the likelihood of detecting a significant change in neurological function. Further, because these are infrequent events, the number of cases associated with specific types of events is small suggesting that a much larger, randomized study will be necessary to determine if the choice of anesthesia impacts the likelihood of detecting a significant change in neural function that would impact outcome. Of note six patients had a slow SSEP amplitude decline which eventually reached alert threshold consistent with the ‘‘fade’’ noted by Lyon [3]. It is also unknown to what extent cases where amplitude decline which did not reach alert criteria (50 % reduction) influenced the variability data in this series. Since amplitude criteria was not used for tcMEP alerts, it is unknown to what extent other cases might have been excluded if amplitude changes had been used. Further studies will be needed to address these issues. Although no new neurological deficits were observed in the patients selected for data analysis it is not possible to exclude that some minor pathology was present in these patients which influenced the data. Further, it is not possible to exclude that the anesthetic technique changed the sensitivity for an alert with minor pathology or events which would lead to a deficit if not corrected. As such the data included in the results give a picture of monitoring in cases where neither monitoring changes reached alert threshold or where new neurological deficits were detected postoperatively. Further study will be necessary to determine if the anesthetic techniques change the sensitivity for impending pathology and neurological outcome. Since the choice of anesthesia was not random, the TIVA group may have preferentially included patients with preexisting neurological or vascular compromise (e.g. diabetes, hypertension) thus biasing the comparison. In addition, if that were the case, the amplitudes of the TIVA group would be expected to be lower such that the difference between the TIVA group and the INHAL group would be minimized. It is not clear what criteria anesthesiologists used for the inclusion of desflurane or not, but at least one practitioner did consciously consider the patients neurological status before utilizing desflurane. We also did not determine the patient population where desflurane produced minimal effects; it is possible that the effects of halogenated anesthetics on the abnormal or injured central nervous system may be more profound than the effects on a normal nervous system. We chose to use average percent trial to trial change to assess variability since this change is used as part of the alert criteria when assessing SSEPs and has been used by some individuals for assessing tcMEPs as well. Coefficient of variation (standard deviation of the trial values divided by the mean) has been used in some studies and we initially calculated the variability using this statistical measure (no
J Clin Monit Comput
differences were seen) [2, 4–6]. However, since our amplitude data is not normally distributed, we chose to use the more intuitive percentage change values. Similar to the amplitude data, this observational study is likely underpowered to detect statistically significant differences between the groups but in these patients there does not appear to be a clinically significant difference in variability. The variability data reveals that patients with large trial to trial average percentage change had lower average amplitudes. This is consistent with the influence of noise where a smaller signal to noise ratio can result in more amplitude variability. This suggests that responses with larger amplitude may have less variability thereby allowing amplitude criteria to be more sensitive to impending neural compromise. Also notable is that even at low average amplitudes, some patients had low variability suggesting individual variability. Thus, patient characteristics, the monitoring technique, anesthesia choice, and other management variables (e.g. blood pressure) could impact on the signal variability. Further study will be necessary to determine if methods to improve the signal to noise ratio may reduce variability in an individual patient. Since the halogenated agents work by different mechanisms than propofol [1], this could explain why in some patients inhalational agents are acceptable for monitoring at low concentrations when propofol infusion is reduced but their use is unacceptable in other patients being monitored when a higher propofol infusion is being used without a halogenated agent. A non-linear dose effect response could also explain why low doses of two drugs may produce similar effects to a high dose of one agent in some patients but not in others. Further studies will be needed to determine if other inhalational agents (sevoflurane, isoflurane, nitrous oxide) behave similar to desflurane in this case series. However, desirable cortical contributions to the anesthesia management (amnesia and unconsciousness) occur at 0.5 MAC or below. Since nitrous oxide is known to be synergistic with inhalational agents in the depression of the SSEP, further studies will also be needed to determine which patients may have acceptable monitoring with desflurane with nitrous oxide [7, 8]. Whereas propofol has its primary action via enhancement of GABA via actions at the GABAa receptor, the halogenated agents such as desflurane act also at the pre and post synaptic NMDA receptor, by action at the nACh, sodium, and potassium K2p channels, by enhancement of glycine inhibition, by action at serotonin type 3 receptors, and by interaction in the hydrophobic region of the cell membrane bilayer on the Na?/K? ATP’ase channel) [9– 15]. All of these actions are thought to contribute to and help insure amnesia and unconsciousness at 0.3–0.5 MAC. These effects explain why an accompanying propofol infusion can be reduced or is not needed.
In addition, the effects of desflurane contribute to antinociception by blocking noxious sensory stimuli which explains the reduction in the sufentanil infusion from 0.36 to 0.26 lg/kg/h. in these patients (data not shown, p = 0.0026). This effect occurs at 0.1–0.5 MAC through actions at the serotonin receptors and by a reduction in sensory transmission through the thalamus at concentrations less than 0.5 MAC [13, 16, 17]. In addition, the halogenated agents are thought to reduce sensory afferent activity by altering the effects of descending signals from the brain that influence the processing of afferent sensory stimuli in the spinal cord [11, 13]. Since these occur via non-opioid mechanisms, these actions may explain why inhalational agents assist in the management of opioid tolerant patients. A non-linear dose effect of inhalational agents has been seen in studies in animals and observations in humans which have shown that the major anesthetic effect on cortical SSEP responses is non-linear and occurs above a narrow concentration in the 0.5-1 % inspired isoflurane at the concentrations where unconsciousness occurs. As such, a low dose of inhalational agent (below 1/2 MAC) may be associated with minimal change in the cortical SSEP and tcMEP in some patients. This ‘‘on–off’’ effect on cortical SSEPs is consistent with study findings showing interference with transmission through the thalamus and general inhibition of cortical sensory processing with their use [9, 18–22]. Finally, in the same fashion as other halogenated agents, desflurane contributes to immobility through actions at the glycine, NMDA, and GABAa receptors within the spinal cord [13, 23–25]. Although desirable clinically to avoid patient movement from surgical stimulation, these effects on the spinal cord motor pathway explain the loss of the ability to acquire tcMEP responses at concentrations above 0.3–0.5 MAC [2, 26–29]. Because desflurane can reduce the signal amplitude, it might lead to more variability. However, low amplitude and high variability can also be seen with pure TIVA suggesting TIVA also may not be an ideal anesthetic in some patients with respect to variability. Further work will be needed to determine if anesthetics which are associated with amplitude enhancement (e.g. ketamine, etomidate) or stimulation techniques associated with signal enhancement (e.g. double train, spatial enhancement) may reduce variability when low amplitude responses would otherwise result with TIVA and conventional stimulation methods. Similarly we do not know the impact of other halogenated agents (e.g. sevoflurane, isoflurane) or nitrous oxide on amplitude or variability when it is used to supplement TIVA. This suggests that the use of desflurane and other inhalational agents might best be reserved for patients with robust tcMEP responses. If an agent is chosen, then
123
J Clin Monit Comput
consideration should be made to choose an insoluble agent (e.g. desflurane, sevoflurane, nitrous oxide) so that it can be readily eliminated if the monitoring indicates that the acquired responses are either absent or too small to be usable such as the three patients identified in the study period. Another limitation of this study is its retrospective nature. Although the assignment of anesthesiologists was not linked to specific patients, the use of halogenated agents was not random. With respect to the propofol infusion rate, this was selected by the anesthesia provider without specific criteria such that a different result might have been observed if an objective endpoint was used. With respect to case selection, the emphasis on cases where the procedure involved the posterior approach to the lumbar spine may suggest that these results may not be as applicable to other approaches such as the cervical spine. It is also unclear if the differences between groups of gender, age, weight, and case duration impact the transferability of these results. Of particular importance is that the patient’s pathology, neural functioning, and presence of preoperative hypnotic and opioid use was not case matched or random. Finally there may have been a preference to use halogenated agents in patients with chronic opioid use and tolerance who may react differently to the addition of low dose halogenated agents.
5 Conclusion This case series demonstrates that balanced anesthesia with 1/2 MAC of desflurane can be used in some patients during spine surgery utilizing SSEP and tcMEP monitoring. However, in certain patients, it may be advisable to avoid their use in favor of a propofol–opioid TIVA. Further studies will be necessary to fully quantitate the effect of using desflurane to determine which patients may derive the benefits of their addition without producing undesirable effects on the ability to monitor. Further studies will also be needed to determine if other inhalational agents also can be used similarly. Conflict of interest of interest.
The authors declare that they have no conflict
References 1. Alkire MT, Hudetz AG, Tononi G. Consciousness and anesthesia. Science. 2008;322(5903):876–80. 2. Woodforth IJ, Hicks RG, Crawford MR, Stephen JP, Burke DJ. Variability of motor-evoked potentials recorded during nitrous oxide anesthesia from the tibialis anterior muscle after transcranial electrical stimulation. Anesth Analg. 1996;82(4):744–9.
123
3. Lyon R, Feiner J, Lieberman JA. Progressive suppression of motor evoked potentials during general anesthesia: the phenomenon of ‘‘anesthetic fade.’’ J Neurosurg Anesthesiol. 2005;17(1): 13–9. 4. Darling WG, Wolf SL, Butler AJ. Variability of motor potentials evoked by transcranial magnetic stimulation depends on muscle activation. Exp Brain Res. 2006;174(2):376–85. 5. Jung NH, Delvendahl I, Kuhnke NG, Hauschke D, Stolle S, Mall V. Navigated transcranial magnetic stimulation does not decrease the variability of motor-evoked potentials. Brain Stimul. 2010;3(2):87–94. 6. Kiers L, Cros D, Chiappa KH, Fang J. Variability of motor potentials evoked by transcranial magnetic stimulation. Electroencephalogr Clin Neurophysiol. 1993;89(6):415–23. 7. Sloan T, Sloan H, Rogers J. Nitrous oxide and isoflurane are synergistic with respect to amplitude and latency effects on sensory evoked potentials. J Clin Monit Comput. 2010;24(2): 113–23. 8. Koitabashi T, Johansen JW, Sebel PS. Remifentanil dose/electroencephalogram bispectral response during combined propofol/ regional anesthesia [see comment]. Anesth Analg. 2002;94(6): 1530–3. 9. Kopp Lugli A, Yost CS, Kindler CH. Anaesthetic mechanisms: update on the challenge of unravelling the mystery of anaesthesia. Eur J Anaesthesiol. 2009;26(10):807–20. 10. Son Y. Molecular mechanisms of general anesthesia. Korean J Anesthesiol. 2010;59(1):3–8. 11. Campagna JA, Miller KW, Forman SA. Mechanisms of actions of inhaled anesthetics. N Engl J Med. 2003;348(21):2110–24. 12. Bonhomme V, Boveroux P, Hans P, Brichant JF, Vanhaudenhuyse A, Boly M, Laureys S. Influence of anesthesia on cerebral blood flow, cerebral metabolic rate, and brain functional connectivity. Curr Opin Anaesthesiol. 2011;24(5):474–9. 13. Rudolph U, Antkowiak B. Molecular and neuronal substrates for general anaesthetics. Nat Rev Neurosci. 2004;5(9):709–20. 14. Jones MV, Harrison NL. Effects of volatile anesthetics on the kinetics of inhibitory postsynaptic currents in cultured rat hippocampal neurons. J Neurophysiol. 1993;70(4):1339–49. 15. Chau PL. New insights into the molecular mechanisms of general anesthetics. Br J Pharmacol. 2010;161:288–307. 16. Chortkoff BS, Eger EI 2nd, Crankshaw DP, Gonsowski CT, Dutton RC, Ionescu P. Concentrations of desflurane and propofol that suppress response to command in humans. Anesth Analg. 1995;81(4):737–43. 17. Dwyer R, Bennett HL, Eger EI 2nd, Heilbron D. Effects of isoflurane and nitrous oxide in subanesthetic concentrations on memory and responsiveness in volunteers. Anesthesiology. 1992; 77(5):888–98. 18. da Costa VV, Saraiva RA, de Almeida AC, Rodrigues MR, Nunes LG, Ferreira JC. The effect of nitrous oxide on the inhibition of somatosensory evoked potentials by sevoflurane in children. Anaesth Intensiv Care. 2001;56:202–7. 19. Manninen PH, Lam AM, Nicholas JF. The effects of isoflurane and isoflurane-nitrous oxide anesthesia on brainstem auditory evoked potentials in humans. Anesth Analg. 1985;64(1):43–7. 20. Samra SK, Vanderzant CW, Domer PA, Sackellares JC. Differential effects of isoflurane on human median nerve somatosensory evoked potentials. Anesthesiology. 1987;66(1):29–35. 21. Hosick EC, Clark DL, Adam N, Rosner BS. Neurophysiological effects of different anesthetics in conscious man. J Appl Physiol. 1971;31(6):892–8. 22. Griffiths R, Norman RI. Effects of anaesthetics on uptake, synthesis and release of transmitters. Br J Anaesth. 1993;71(1):96–107. 23. Stabernack C, Sonner JM, Laster M, Zhang Y, Xing Y, Sharma M, Eger EI 2nd. Spinal N-methyl-d-aspartate receptors may contribute to the immobilizing action of isoflurane. Anesth Analg. 2003;96(1):102–7.
J Clin Monit Comput 24. Zhang Y, Laster MJ, Hara K, Harris RA, Eger EI 2nd, Stabernack CR, Sonner JM. Glycine receptors mediate part of the immobility produced by inhaled anesthetics. Anesth Analg. 2003;96(1):97–101. 25. Mashour GA, Forman SA, Campagna JA. Mechanisms of general anesthesia: from molecules to mind. Best Pract Res Clin Anaesthesiol. 2005;19(3):349–64. 26. Haghighi S, Madsen R, Green K. Suppression of motor evoked potentials by inhalation anesthetics. J Neurosurg Anesthes. 1990; 2:73–6.
27. Haghighi SS, Green KD, Oro JJ, Drake RK, Kracke GR. Depressive effect of isoflurane anesthesia on motor evoked potentials. Neurosurgery. 1990;26(6):993–7. 28. Loughnan BA, Anderson SK, Hetreed MA, Weston PF, Boyd SG, Hall GM. Effects of halothane on motor evoked potential recorded in the extradural space. Br J Anaesth. 1989;63(5):561–4. 29. Zentner J, Albrecht T, Heuser D. Influence of halothane, enflurane, and isoflurane on motor evoked potentials. Neurosurgery. 1992;31(2):298–305.
123
ORIGINAL PAPER
Intraoperative neurophysiological monitoring during spine surgery with total intravenous anesthesia or balanced anesthesia with 3 % desflurane Tod B. Sloan • J. Richard Toleikis Sandra C. Toleikis • Antoun Koht
•
Received: 14 September 2013 / Accepted: 12 March 2014 Ó Springer Science+Business Media New York 2014
Abstract Total intravenous anesthesia (TIVA) with propofol and opioids is frequently utilized for spinal surgery when somatosensory evoked potentials (SSEPs) and transcranial motor evoked potentials (tcMEPs) are monitored. Many anesthesiologists would prefer to utilize low dose halogenated anesthetics (e.g. 1/2 MAC). We examined our recent experience using 3 % desflurane or TIVA during spine surgery to determine the impact on propofol usage and on the evoked potential responses. After institutional review board approval we conducted a retrospective review of a 6 month period for adult spine patients who were monitored with SSEPs and tcMEPs. Cases were included for the study if anesthesia was conducted with propofol–opioid TIVA or 3 % desflurane supplemented with propofol or opioid infusions as needed. We evaluated the propofol infusion rate, cortical amplitudes of the SSEPs (median nerve, posterior tibial nerve), amplitudes and stimulation voltage for eliciting the tcMEPs (adductor pollicis brevis, tibialis anterior) and the amplitude variability of the SSEP and tcMEP responses
T. B. Sloan (&) Department of Anesthesiology, Anschutz Office West (AO1), MS 8202, University of Colorado Denver School of Medicine, 12631 E 17th Avenue, Aurora, CO 80045, USA e-mail: [email protected] J. R. Toleikis Department of Anesthesiology, Rush Medical College, Rush University Medical Center, Chicago, IL, USA S. C. Toleikis Department of Anesthesiology, Rush University Medical Center, Chicago, IL, USA A. Koht Departments of Anesthesiology, Neurosurgery, and Neurology, Northwestern University, Chicago, IL, USA
as assessed by the average percentage trial to trial change. Of the 156 spine cases included in the study, 95 had TIVA with propofol–opioid (TIVA) and 61 had 3 % expired desflurane (INHAL). Three INHAL cases were excluded because the desflurane was eliminated because of inadequate responses and 26 cases (16 TIVA and 10 INHAL) were excluded due to significant changes during monitoring. Propofol infusion rates in the INHAL group were reduced from the TIVA group (average 115–45 lg/kg/min) (p \ 0.00001) with 21 cases where propofol was not used. No statistically significant differences in cortical SSEP or tcMEP amplitudes, tcMEP stimulation voltages nor in the average trial to trial amplitude variability were seen. The data from these cases indicates that 1/2 MAC (3 %) desflurane can be used in conjunction with SSEP and tcMEP monitoring for some adult patients undergoing spine surgery. Further studies are needed to confirm the relative benefits versus negative effects of the use of desflurane and other halogenated agents for anesthesia during procedures on neurophysiological monitoring involving tcMEPs. Further studies are also needed to characterize which patients may or may not be candidates for supplementation such as those with neural dysfunction or who are opioid tolerant from chronic use. Keywords Propofol Desflurane Total intravenous anesthesia Somatosensory evoked potentials Motor evoked potentials Spinal surgery
1 Introduction Total intravenous anesthesia (TIVA) is often utilized when electrophysiological monitoring is used during spine surgery. This is frequently accomplished using infusions of propofol and an opioid (e.g. sufentanil). Anesthesiologists
123
J Clin Monit Comput
often would prefer to supplement this with low dose (e.g. 1/2 MAC) of a halogenated anesthetic agent (e.g. desflurane) in these cases because of their contribution to the goals of anesthesia and as well as a means to conserve propofol. The use of an inhalational agent such as desflurane with intravenous anesthetics may be particularly helpful in patients with opioid tolerance since the halogenated agents work through synaptic mechanisms different than the opioids [1]. Because both propofol and halogenated agents have been associated with amplitude reduction of cortical somatosensory evoked potentials (SSEPs) and muscle recordings from transcranially evoked motor evoked potentials (tcMEPs), the comparative impact on monitoring needs to be characterized. To our knowledge no systematic study of propofol–opioid anesthesia versus 1/2 MAC of desflurane with propofol and/or opioid infusions has been published. The potential impact of inhalational agents on tcMEPs has been specifically raised by Woodforth who indicated tcMEP responses may be ‘‘difficult to rely on’’ because of marked variations in their amplitude when nitrous oxide and isoflurane were used [2]. Hence the variability of the responses needs to be characterized when inhalational agents are added since this may change the usefulness of amplitude criteria during monitoring. We retrospectively reviewed our experience with spine surgery cases when propofol–opioid anesthesia and 1/2 MAC desflurane with propofol and/or opioid infusions were utilized in order to determine the impact of the halogenated agent on cortical SSEPs, multipulse tcMEP recordings, and the utilized dose of propofol.
2 Methods After Institutional Review Board approval, the intraoperative monitoring records for a 6 month period between September 1, 2012 and March 31, 2013 were reviewed to identify spinal surgeries when intraoperative monitoring was utilized. The anesthesia records of these cases were used to identify procedures when the anesthesia management utilized only propofol–opioid infusions (TIVA group) or 3 % desflurane with infusions of propofol and/or opioid (INHAL group). The infusion rates of propofol during the majority of the surgical period were recorded from the electronic medical record. We excluded cases where other intravenous agents (e.g. dexmeditomidine, ketamine, and lidocaine) or other inhalational agents (e.g. sevoflurane, nitrous oxide) were used. Other recorded data included the surgical procedure, case duration, patient gender, age, and weight. During the selected investigative period, several individuals provided anesthesia services. No particular assignment schedule was followed and the use of halogenated
123
agents was decided upon by the individual providers. When desflurane was used, it was routinely utilized during the procedure at 3 % expired (1/2 minimal alveolar concentration: MAC) with infusions of propofol and/or opioids as needed. Cases where desflurane was used briefly at induction and eliminated prior to monitoring were considered for inclusion in the TIVA group. Since it is known that inhalational agents (especially when accompanied with nitrous oxide) can markedly depress the amplitude or ability to acquire SSEP or tcMEP responses, those cases where we were unable to acquire responses in the presence of desflurane and the desflurane was terminated for monitoring were not included in the case series. The rates of propofol infusion were adjusted clinically at the discretion of the anesthesiologist as were the choices and rates of opioid infusion (sufentanil or remifentanil). In these cases electrophysiological monitoring was conducted using a Cadwell Cascade Elite (Cadwell Laboratories, Inc., Kennewick, WA). Cortical SSEPs were recorded with sterile subdermal needles placed at C30 , C40 , Cz0 and referenced to Fz using bandpass filtration of 30-500 Hz (all electrode positions based on the International 10–20 system). Each SSEP averaged response consisted of two hundred, 50 or 100 ms traces of cortical activity that resulted from stimulation of a median nerve (MN) at the wrist or posterior tibial nerve (PTN) at the ankle, respectively. Stimulation was provided through sterile subdermal needles at a frequency of 2.66 Hertz (Hz) using 100 ms pulses at a current intensity just above motor threshold. The cortical amplitudes of the SSEPs that resulted from median and posterior tibial nerve stimulation were recorded for each case. TcMEP responses were recorded from sterile subdermal needles placed in the abductor pollicis brevis and tibialis anterior muscles using bandpass filtration of 10–3,000 Hz. These responses were elicited by transcranial electrical stimulation at C1–C2 using a train of six pulses with a 50 ms pulse width and an interpulse interval of 2 ms. The stimulation voltage was chosen by the monitoring professional based on the intensity that was needed for response acquisition. For the purpose of analysis of the effect of the anesthesia on the SSEP amplitudes, the last monitored response prior to stopping or tapering the propofol and opioid infusions in anticipation of the conclusion of the procedure was used. When left and right-sided responses differed in amplitude, the response with the largest amplitude was chosen. Similar to the analysis of the SSEP amplitudes, the amplitudes of the last responses from the abductor pollicis brevis (APB) and tibialis anterior (TA) prior to terminal anesthetic change were recorded. Similar to that of the SSEPs, when the left and right-sided responses differed in amplitude, the response with the largest amplitude was selected. The
J Clin Monit Comput
stimulation voltages were also recorded at that time. If more than one stimulation intensity was used, the highest one was recorded. In order to quantitate the effect of the anesthesia on amplitude variability, the data were grouped into data sets for each patient limb. Hence, for each patient, separate SSEP data sets were created using all of the evoked response recordings for the left and right median and posterior tibial nerves (4 data sets per patient). Similarly, four additional data sets were created using the tcMEP amplitudes of the left and right adductor pollicis brevis and tibialis anterior muscles. For each data set, the average signal variability was calculated by determining the average of the absolute percentage change from each trial to the subsequent response. The median amplitude for each data set was also calculated to characterize the relationship of response variability and amplitude. Data sets where a monitoring alert occurred (e.g. a sustained 50 % or greater drop in the amplitude of any cortical SSEP response or a total loss of tcMEP responses) in conjunction with an operative event (e.g. surgical event, anesthetic bolus, etc.) were excluded from inclusion in the analysis. Chi square analysis was used to determine if there was a significant difference in the occurrence of alerts between the groups. The results were divided into two groups depending on the use of desflurane. The TIVA group utilized only propofol and opioid infusions (sufentanil or remifentanil) whereas for the INHAL group, desflurane was used throughout the case with infusions of propofol and/or opioid as determined by the anesthesiologist. The SSEP and tcMEP data are displayed using box plots because the values are not normally distributed. The box displays the middle half of the data (between the 25th and 75th percentile). The bar within the box represents the median of the data and the vertical line extends between the minimum and maximum values. A student’s t test was used to compare groups. p \ 0.05 was considered significant.
3 Results During the study period, 156 patients were identified for inclusion. Anesthesia was conducted using propofol–opioid infusions (TIVA, 95) or 3 % desflurane with propofol and/or
opioid infusions (INHAL, 61). Three cases were excluded from the INHAL group because desflurane was removed because of inadequate responses (see below) and 26 cases were excluded (TIVA 16, INHAL 10) because of monitoring alerts. The final case series included 127 cases (TIVA 79, INHAL 48) which included 15 anterior cervical procedures (TIVA 13, INHAL 2), 32 posterior cervical procedures (TIVA 23, INHAL 9), and 82 posterior thoracic and/or lumbar procedures (TIVA 43, INHAL 37). Table 1 shows the characteristics of the groups and the average propofol infusion rates. The difference between the propofol infusion rates was statistically significant (p \ 0.0001). Three patients in the INHAL group were excluded because the desflurane was terminated when the monitored responses were not discernable from the background noise. In one case there were no lower extremity SSEP responses in the presence of desflurane but the responses returned (average amplitude 0.3 lV) after termination of desflurane usage. In the second case, one lower extremity tcMEP was absent (the other limb response had an amplitude less than 50 lV). Bilateral responses were present (average amplitude 400 lV) when the use of desflurane was terminated. In the third case no lower limb tcMEP responses could be recorded in the presence of desflurane and the responses were still not discernable after desflurane was eliminated. Thirty-four intraoperative monitoring alerts occurred in 26 patients (TIVA 16, INHAL 10) which were excluded from analysis. Five alerts (TIVA 2, INHAL 3) occurred in five patients who had reversible amplitude decline or loss in the left leg response associated with occlusion of the left iliac artery during anterior exposure to the spine. Seven alerts (TIVA 4, INHAL 3) occurred in seven patients who had amplitude decline of one or both upper extremity SSEP responses during surgery on the thoracic or lumbar spine which was partially corrected by repositioning the arms. Six alerts (TIVA 3, INHAL 3) occurred in six patients where the SSEP amplitude decline occurred gradually over several hours eventually reaching 50 % or less. Fifteen alerts (TIVA 13, INHAL 2) were associated with surgical events in 12 patients where abrupt decline or loss of responses was associated with manipulation of a spinal cord tumor (5), hematoma removal (1), spinal cord traction (3), pedicle screw placement (2), dural tear (1), neck positioning (2), and corpectomy (1). One alert resulted from administration of neuromuscular blockade. The
Table 1 Basic case characteristics Gender (number M, F)
Age (years) (Average, min, max)
Weight (kg) (Average, SD)
Duration (min) (Average, SD)
Propofol (lg/kg/min) (Average, SD)
TIVA
41, 38
56 (22, 87)
79 (17)
345 (142)
115 (24)
INHAL
21, 27
59 (24, 83)
81 (24)
330 (123)
49 (49)
123
J Clin Monit Comput
occurrence of total alerts between groups was not statistically significant (p [ 0.94). Assessment of postoperative outcome revealed no new neurological deficits in any of the patients included in the data analysis. In addition, in the patients with alerts, no new neurological deficits were observed except in the four patients with the five alerts related to manipulation of spinal cord tumors (TIVA 3, INHAL 1) and hematoma removal (INHAL). The number of cases is too small to determine if the presence of the inhalational agent changed the sensitivity of the monitoring to the impending pathology.
Fig. 1 Median nerve and posterior tibial nerve cortical SSEP amplitudes. Box plot of SSEP cortical amplitudes from stimulation of the median nerve (MN, left) or posterior tibial nerve (PTN, right) where anesthesia was propofol–opioid TIVA (TIVA) or 3 % desflurane supplemented as needed with propofol and opioid infusions (INHAL). The maximum values are noted if they exceed the limits of the plot Fig. 2 Muscle amplitudes from transcranial motor evoked potential (tcMEP) stimulation. Box plot of tcMEP amplitudes from recordings from the adductor pollicis brevis (APB, left) or tibialis anterior (TA, right) where anesthesia was propofol–opioid TIVA (TIVA, left) or 3 % desflurane supplemented as needed with propofol and opioid infusions (INHAL, right). The maximum values are noted if they exceed the limits of the plot
123
Figure 1 shows the box plots of the amplitudes of the cortical SSEPs that resulted from median and posterior tibial nerve stimulation. The median values for the median nerve SSEP was 1.92 lV (TIVA) and 1.60 lV (INHAL) (p = 0.079). For the posterior tibial nerve the median values were 1.09 lV (TIVA) and 1.00 lV (INHAL) (p = 0.44). Figure 2 shows the box plots of the amplitudes of the tcMEP responses recorded from the APB and TA muscles. The median values for the APB was 2016 lV (TIVA) and 2,867 lV (INHAL) (p = 0.15) and for the TA was 812 lV (TIVA) and 749 lV (INHAL) (p = 0.82). Figure 3 shows the box plots of the stimulation intensities used for eliciting the tcMEP responses (median values 400 V for each group) (p = 0.10). No significant differences were detected between these groups. The results of the trial to trial variability are shown in Fig. 4. The median nerve and posterior tibial nerve cortical SSEP responses revealed an average trial to trial percentage change of 14–18 % with small but not statistically significant differences between groups (MN, p = 0.27) (PTN, p = 0.054). Average variability for individual limbs ranged from 2.9–42.3 % with larger values of variability occurring when the average amplitude of the responses was less than 4 lV and particularly when it was less than 2.5 lV (Fig. 5). Similarly, the amplitudes of tcMEP responses demonstrated an average 34–41 % inter-trial change in amplitude for which small but not statistically significant differences between groups (Fig. 6) were found (APB, p = 0.80) (TA, p = 0.66). Average variability for tcMEP responses from individual limbs ranged from 3 to over 300 % with larger variability in response amplitude occurring when the average amplitude was less than 4,500 lV and particularly when it was below 2,000 lV (Fig. 7).
J Clin Monit Comput
Fig. 5 Average cortical SSEP amplitude trial to trial percentage change versus average limb amplitude. Plot of average trial to trial percentage change of SSEP cortical amplitudes (MN and PTN combined) versus average amplitude where anesthesia was propofol– opioid TIVA (TIVA) or propofol–opioid-3 % desflurane (INHAL) Fig. 3 Transcranial motor evoked potential stimulation voltage. Box plot of tcMEP stimulation voltage used in the case series where anesthesia was propofol–opioid TIVA (TIVA, left) or 3 % desflurane supplemented as needed with propofol and opioid infusions (INHAL, right)
Fig. 4 Average trial to trial percentage change of cortical SSEP amplitude. Box plot of average trial to trial percentage change of SSEP cortical amplitudes from stimulation of the median nerve (MN, left) or posterior tibial nerve (PTN, right) where anesthesia was propofol–opioid TIVA (TIVA) or 3 % desflurane supplemented as needed with propofol and opioid infusions (INHAL). The maximum values are noted if they exceed the limits of the plot
4 Discussion These data demonstrate that 3 % desflurane can be used in some patients during the monitoring of cortical SSEPs and the muscle responses of tcMEPs. The cases in this series show that, as a group, the amplitudes of the cortical SSEPs, amplitudes of the tcMEP muscle responses, and the tcMEP
Fig. 6 Average trial to trial percentage change of motor evoked potential amplitude. Box plot of average trial to trial percentage change of tcMEP amplitudes from recordings from the adductor pollicis brevis (APB, left) or tibialis anterior (TA, right) where anesthesia was propofol–opioid TIVA (TIVA, left) or propofol– opioid-3 % desflurane (INHAL, right). The maximum values are noted if they exceed the limits of the plot
stimulation voltages are comparable. Further, as a group, the variability of the responses as measured by the average percentage change from trial to trial is also comparable. One explanation may be that the depression of the responses associated with 3 % desflurane is offset by the effect of the lower propofol infusions. Certainly this effect would be better characterized in a crossover study designed to compare the effects of the two agents. It should be noted that this observational study is likely underpowered to detect a significant difference between these groups but the
123
J Clin Monit Comput
Fig. 7 Average tcMEP amplitude trial to trial percentage change versus average limb amplitude. Plot of average trial to trial percentage change of tcMEP amplitudes (both APB and TA plotted) versus average amplitude where anesthesia was propofol–opioid TIVA (TIVA) or propofol–opioid-3 % desflurane (INHAL)
data suggests a lack of clinically significant difference in these patients. This is not the case in all patients since three patients had no recordable SSEP or tcMEP responses when 3 % desflurane was initially used and the anesthesia was switched to TIVA. With this switch the responses improved in two cases, particularly the tcMEP responses, suggesting that in their case the effect of desflurane exceeded the impact of a larger dose of propofol. Each of these patients had neurological symptoms (numbness, tingling in an extremity) which may have impacted on the relative anesthetic effect of desflurane and propofol. In one case responses could not be recorded without desflurane consistent with the observation that not all patients have recordable responses under TIVA despite clinical sensory and motor functional deficits. This underscores one limitation of this study since the patients where desflurane was used were not randomly chosen. Practitioners in this case series may have reserved desflurane usage for patients who did not have neural symptoms or a disease process which is associated with more difficult monitoring (numbness, tingling, weakness, diabetes, vascular disease), or where chronic opioid use was known to make anesthesia management more difficult. This suggests that there might be a clinically significant difference in the effect of the choice of anesthesia is some patients. In this case series 26 patients were excluded from analysis of the propofol infusion rate and the response variability because one or more monitoring alerts occurred during the case. Although the frequency of alerts were not different between the groups it is unknown if the type of
123
anesthesia changed the likelihood of detecting a significant change in neurological function. Further, because these are infrequent events, the number of cases associated with specific types of events is small suggesting that a much larger, randomized study will be necessary to determine if the choice of anesthesia impacts the likelihood of detecting a significant change in neural function that would impact outcome. Of note six patients had a slow SSEP amplitude decline which eventually reached alert threshold consistent with the ‘‘fade’’ noted by Lyon [3]. It is also unknown to what extent cases where amplitude decline which did not reach alert criteria (50 % reduction) influenced the variability data in this series. Since amplitude criteria was not used for tcMEP alerts, it is unknown to what extent other cases might have been excluded if amplitude changes had been used. Further studies will be needed to address these issues. Although no new neurological deficits were observed in the patients selected for data analysis it is not possible to exclude that some minor pathology was present in these patients which influenced the data. Further, it is not possible to exclude that the anesthetic technique changed the sensitivity for an alert with minor pathology or events which would lead to a deficit if not corrected. As such the data included in the results give a picture of monitoring in cases where neither monitoring changes reached alert threshold or where new neurological deficits were detected postoperatively. Further study will be necessary to determine if the anesthetic techniques change the sensitivity for impending pathology and neurological outcome. Since the choice of anesthesia was not random, the TIVA group may have preferentially included patients with preexisting neurological or vascular compromise (e.g. diabetes, hypertension) thus biasing the comparison. In addition, if that were the case, the amplitudes of the TIVA group would be expected to be lower such that the difference between the TIVA group and the INHAL group would be minimized. It is not clear what criteria anesthesiologists used for the inclusion of desflurane or not, but at least one practitioner did consciously consider the patients neurological status before utilizing desflurane. We also did not determine the patient population where desflurane produced minimal effects; it is possible that the effects of halogenated anesthetics on the abnormal or injured central nervous system may be more profound than the effects on a normal nervous system. We chose to use average percent trial to trial change to assess variability since this change is used as part of the alert criteria when assessing SSEPs and has been used by some individuals for assessing tcMEPs as well. Coefficient of variation (standard deviation of the trial values divided by the mean) has been used in some studies and we initially calculated the variability using this statistical measure (no
J Clin Monit Comput
differences were seen) [2, 4–6]. However, since our amplitude data is not normally distributed, we chose to use the more intuitive percentage change values. Similar to the amplitude data, this observational study is likely underpowered to detect statistically significant differences between the groups but in these patients there does not appear to be a clinically significant difference in variability. The variability data reveals that patients with large trial to trial average percentage change had lower average amplitudes. This is consistent with the influence of noise where a smaller signal to noise ratio can result in more amplitude variability. This suggests that responses with larger amplitude may have less variability thereby allowing amplitude criteria to be more sensitive to impending neural compromise. Also notable is that even at low average amplitudes, some patients had low variability suggesting individual variability. Thus, patient characteristics, the monitoring technique, anesthesia choice, and other management variables (e.g. blood pressure) could impact on the signal variability. Further study will be necessary to determine if methods to improve the signal to noise ratio may reduce variability in an individual patient. Since the halogenated agents work by different mechanisms than propofol [1], this could explain why in some patients inhalational agents are acceptable for monitoring at low concentrations when propofol infusion is reduced but their use is unacceptable in other patients being monitored when a higher propofol infusion is being used without a halogenated agent. A non-linear dose effect response could also explain why low doses of two drugs may produce similar effects to a high dose of one agent in some patients but not in others. Further studies will be needed to determine if other inhalational agents (sevoflurane, isoflurane, nitrous oxide) behave similar to desflurane in this case series. However, desirable cortical contributions to the anesthesia management (amnesia and unconsciousness) occur at 0.5 MAC or below. Since nitrous oxide is known to be synergistic with inhalational agents in the depression of the SSEP, further studies will also be needed to determine which patients may have acceptable monitoring with desflurane with nitrous oxide [7, 8]. Whereas propofol has its primary action via enhancement of GABA via actions at the GABAa receptor, the halogenated agents such as desflurane act also at the pre and post synaptic NMDA receptor, by action at the nACh, sodium, and potassium K2p channels, by enhancement of glycine inhibition, by action at serotonin type 3 receptors, and by interaction in the hydrophobic region of the cell membrane bilayer on the Na?/K? ATP’ase channel) [9– 15]. All of these actions are thought to contribute to and help insure amnesia and unconsciousness at 0.3–0.5 MAC. These effects explain why an accompanying propofol infusion can be reduced or is not needed.
In addition, the effects of desflurane contribute to antinociception by blocking noxious sensory stimuli which explains the reduction in the sufentanil infusion from 0.36 to 0.26 lg/kg/h. in these patients (data not shown, p = 0.0026). This effect occurs at 0.1–0.5 MAC through actions at the serotonin receptors and by a reduction in sensory transmission through the thalamus at concentrations less than 0.5 MAC [13, 16, 17]. In addition, the halogenated agents are thought to reduce sensory afferent activity by altering the effects of descending signals from the brain that influence the processing of afferent sensory stimuli in the spinal cord [11, 13]. Since these occur via non-opioid mechanisms, these actions may explain why inhalational agents assist in the management of opioid tolerant patients. A non-linear dose effect of inhalational agents has been seen in studies in animals and observations in humans which have shown that the major anesthetic effect on cortical SSEP responses is non-linear and occurs above a narrow concentration in the 0.5-1 % inspired isoflurane at the concentrations where unconsciousness occurs. As such, a low dose of inhalational agent (below 1/2 MAC) may be associated with minimal change in the cortical SSEP and tcMEP in some patients. This ‘‘on–off’’ effect on cortical SSEPs is consistent with study findings showing interference with transmission through the thalamus and general inhibition of cortical sensory processing with their use [9, 18–22]. Finally, in the same fashion as other halogenated agents, desflurane contributes to immobility through actions at the glycine, NMDA, and GABAa receptors within the spinal cord [13, 23–25]. Although desirable clinically to avoid patient movement from surgical stimulation, these effects on the spinal cord motor pathway explain the loss of the ability to acquire tcMEP responses at concentrations above 0.3–0.5 MAC [2, 26–29]. Because desflurane can reduce the signal amplitude, it might lead to more variability. However, low amplitude and high variability can also be seen with pure TIVA suggesting TIVA also may not be an ideal anesthetic in some patients with respect to variability. Further work will be needed to determine if anesthetics which are associated with amplitude enhancement (e.g. ketamine, etomidate) or stimulation techniques associated with signal enhancement (e.g. double train, spatial enhancement) may reduce variability when low amplitude responses would otherwise result with TIVA and conventional stimulation methods. Similarly we do not know the impact of other halogenated agents (e.g. sevoflurane, isoflurane) or nitrous oxide on amplitude or variability when it is used to supplement TIVA. This suggests that the use of desflurane and other inhalational agents might best be reserved for patients with robust tcMEP responses. If an agent is chosen, then
123
J Clin Monit Comput
consideration should be made to choose an insoluble agent (e.g. desflurane, sevoflurane, nitrous oxide) so that it can be readily eliminated if the monitoring indicates that the acquired responses are either absent or too small to be usable such as the three patients identified in the study period. Another limitation of this study is its retrospective nature. Although the assignment of anesthesiologists was not linked to specific patients, the use of halogenated agents was not random. With respect to the propofol infusion rate, this was selected by the anesthesia provider without specific criteria such that a different result might have been observed if an objective endpoint was used. With respect to case selection, the emphasis on cases where the procedure involved the posterior approach to the lumbar spine may suggest that these results may not be as applicable to other approaches such as the cervical spine. It is also unclear if the differences between groups of gender, age, weight, and case duration impact the transferability of these results. Of particular importance is that the patient’s pathology, neural functioning, and presence of preoperative hypnotic and opioid use was not case matched or random. Finally there may have been a preference to use halogenated agents in patients with chronic opioid use and tolerance who may react differently to the addition of low dose halogenated agents.
5 Conclusion This case series demonstrates that balanced anesthesia with 1/2 MAC of desflurane can be used in some patients during spine surgery utilizing SSEP and tcMEP monitoring. However, in certain patients, it may be advisable to avoid their use in favor of a propofol–opioid TIVA. Further studies will be necessary to fully quantitate the effect of using desflurane to determine which patients may derive the benefits of their addition without producing undesirable effects on the ability to monitor. Further studies will also be needed to determine if other inhalational agents also can be used similarly. Conflict of interest of interest.
The authors declare that they have no conflict
References 1. Alkire MT, Hudetz AG, Tononi G. Consciousness and anesthesia. Science. 2008;322(5903):876–80. 2. Woodforth IJ, Hicks RG, Crawford MR, Stephen JP, Burke DJ. Variability of motor-evoked potentials recorded during nitrous oxide anesthesia from the tibialis anterior muscle after transcranial electrical stimulation. Anesth Analg. 1996;82(4):744–9.
123
3. Lyon R, Feiner J, Lieberman JA. Progressive suppression of motor evoked potentials during general anesthesia: the phenomenon of ‘‘anesthetic fade.’’ J Neurosurg Anesthesiol. 2005;17(1): 13–9. 4. Darling WG, Wolf SL, Butler AJ. Variability of motor potentials evoked by transcranial magnetic stimulation depends on muscle activation. Exp Brain Res. 2006;174(2):376–85. 5. Jung NH, Delvendahl I, Kuhnke NG, Hauschke D, Stolle S, Mall V. Navigated transcranial magnetic stimulation does not decrease the variability of motor-evoked potentials. Brain Stimul. 2010;3(2):87–94. 6. Kiers L, Cros D, Chiappa KH, Fang J. Variability of motor potentials evoked by transcranial magnetic stimulation. Electroencephalogr Clin Neurophysiol. 1993;89(6):415–23. 7. Sloan T, Sloan H, Rogers J. Nitrous oxide and isoflurane are synergistic with respect to amplitude and latency effects on sensory evoked potentials. J Clin Monit Comput. 2010;24(2): 113–23. 8. Koitabashi T, Johansen JW, Sebel PS. Remifentanil dose/electroencephalogram bispectral response during combined propofol/ regional anesthesia [see comment]. Anesth Analg. 2002;94(6): 1530–3. 9. Kopp Lugli A, Yost CS, Kindler CH. Anaesthetic mechanisms: update on the challenge of unravelling the mystery of anaesthesia. Eur J Anaesthesiol. 2009;26(10):807–20. 10. Son Y. Molecular mechanisms of general anesthesia. Korean J Anesthesiol. 2010;59(1):3–8. 11. Campagna JA, Miller KW, Forman SA. Mechanisms of actions of inhaled anesthetics. N Engl J Med. 2003;348(21):2110–24. 12. Bonhomme V, Boveroux P, Hans P, Brichant JF, Vanhaudenhuyse A, Boly M, Laureys S. Influence of anesthesia on cerebral blood flow, cerebral metabolic rate, and brain functional connectivity. Curr Opin Anaesthesiol. 2011;24(5):474–9. 13. Rudolph U, Antkowiak B. Molecular and neuronal substrates for general anaesthetics. Nat Rev Neurosci. 2004;5(9):709–20. 14. Jones MV, Harrison NL. Effects of volatile anesthetics on the kinetics of inhibitory postsynaptic currents in cultured rat hippocampal neurons. J Neurophysiol. 1993;70(4):1339–49. 15. Chau PL. New insights into the molecular mechanisms of general anesthetics. Br J Pharmacol. 2010;161:288–307. 16. Chortkoff BS, Eger EI 2nd, Crankshaw DP, Gonsowski CT, Dutton RC, Ionescu P. Concentrations of desflurane and propofol that suppress response to command in humans. Anesth Analg. 1995;81(4):737–43. 17. Dwyer R, Bennett HL, Eger EI 2nd, Heilbron D. Effects of isoflurane and nitrous oxide in subanesthetic concentrations on memory and responsiveness in volunteers. Anesthesiology. 1992; 77(5):888–98. 18. da Costa VV, Saraiva RA, de Almeida AC, Rodrigues MR, Nunes LG, Ferreira JC. The effect of nitrous oxide on the inhibition of somatosensory evoked potentials by sevoflurane in children. Anaesth Intensiv Care. 2001;56:202–7. 19. Manninen PH, Lam AM, Nicholas JF. The effects of isoflurane and isoflurane-nitrous oxide anesthesia on brainstem auditory evoked potentials in humans. Anesth Analg. 1985;64(1):43–7. 20. Samra SK, Vanderzant CW, Domer PA, Sackellares JC. Differential effects of isoflurane on human median nerve somatosensory evoked potentials. Anesthesiology. 1987;66(1):29–35. 21. Hosick EC, Clark DL, Adam N, Rosner BS. Neurophysiological effects of different anesthetics in conscious man. J Appl Physiol. 1971;31(6):892–8. 22. Griffiths R, Norman RI. Effects of anaesthetics on uptake, synthesis and release of transmitters. Br J Anaesth. 1993;71(1):96–107. 23. Stabernack C, Sonner JM, Laster M, Zhang Y, Xing Y, Sharma M, Eger EI 2nd. Spinal N-methyl-d-aspartate receptors may contribute to the immobilizing action of isoflurane. Anesth Analg. 2003;96(1):102–7.
J Clin Monit Comput 24. Zhang Y, Laster MJ, Hara K, Harris RA, Eger EI 2nd, Stabernack CR, Sonner JM. Glycine receptors mediate part of the immobility produced by inhaled anesthetics. Anesth Analg. 2003;96(1):97–101. 25. Mashour GA, Forman SA, Campagna JA. Mechanisms of general anesthesia: from molecules to mind. Best Pract Res Clin Anaesthesiol. 2005;19(3):349–64. 26. Haghighi S, Madsen R, Green K. Suppression of motor evoked potentials by inhalation anesthetics. J Neurosurg Anesthes. 1990; 2:73–6.
27. Haghighi SS, Green KD, Oro JJ, Drake RK, Kracke GR. Depressive effect of isoflurane anesthesia on motor evoked potentials. Neurosurgery. 1990;26(6):993–7. 28. Loughnan BA, Anderson SK, Hetreed MA, Weston PF, Boyd SG, Hall GM. Effects of halothane on motor evoked potential recorded in the extradural space. Br J Anaesth. 1989;63(5):561–4. 29. Zentner J, Albrecht T, Heuser D. Influence of halothane, enflurane, and isoflurane on motor evoked potentials. Neurosurgery. 1992;31(2):298–305.
123
