Somatosensory evoked potentials (SSEPs) are often used to monitor the spinal cord during scoliosis surgery (1). Stimulating the posterior tibial nerves and recording the potentials evoked in the cortical and subcortical regions by using scalp electrodes generates the monitored SSEP (2).
Inhaled anesthetics, given to ensure unconsciousness, adversely affect the evoked cortical response of the SSEP (3–7). This may obscure the SSEP, by depression and desynchronization of the wave-form amplitude, so that meaningful data about spinal cord function cannot be obtained. In electrically noisy operating room environments, a low-amplitude SSEP (i.e., a low signal-to-noise ratio) may also make it difficult to measure the latency of the SSEP within a clinically appropriate averaging time.
In recent years, an end-organ measure of sedation has been developed and validated. This is the measurement of the bispectral index (BIS™). A BIS value of 60 is believed to reflect depression of brain function adequate to ensure unconsciousness (8–10).
Use of BIS allows titration of medication to the same degree of unconsciousness in each patient and eliminates the guessing previously used to judge depth of anesthesia. This provides standardization of the neurobiological effect of the anesthetic in an individual patient. With this tool, it is possible to compare the effects of different anesthetics on SSEPs at comparable sedation depths as measured by BIS. In this randomized study, we compared the relative effect of isoflurane and desflurane at a predetermined BIS value of 60 on the posterior tibial nerve-generated cortical SSEP.
This study was approved by the hospital IRB, and informed written consent/assent was obtained from each parent/subject as required by the IRB. Patients were eligible for admission to the study if they were between 4 and 18 yr of age, were ASA status I–III, and were undergoing elective correction of idiopathic scoliosis under general anesthesia with the use of SSEP monitoring of the spinal cord. Patients were excluded if neurological abnormalities were present. The study period was after spinal distraction and at a time the surgeon determined that SSEP monitoring was no longer indicated for patient care.
Patients were randomized to receive either isoflurane (“ISO first” group) or desflurane (“DES first” group) as their first inhaled anesthetic. This first anesthetic was used throughout surgery and until completion of the first two study SSEP measurements. It was then discontinued, and the second inhaled anesthetic was given until completion of the study.
Patients were premedicated with midazolam if clinically indicated. The induction of anesthesia was with IV propofol or inhaled sevoflurane. Once the patient was asleep, an IV was started, if not already present, and a bolus of fentanyl 3 μg/kg was given, together with a neuromuscular blocking drug (either vecuronium or cisatracurium). Endotracheal intubation was performed, and anesthesia was maintained by intermittent positive pressure ventilation to normocapnia with air, oxygen, and either isoflurane or desflurane, along with an infusion of fentanyl 1.5–2 μg · kg−1 · h−1 (the infusion rate of fentanyl was clinically determined within the first hour after induction, according to the patient’s clinical response to surgery). The choice of isoflurane or desflurane as the first anesthetic was determined by prior randomization. No nitrous oxide was used. Toward the end of surgery, before the study period commenced, a wake-up test was performed. This involved reversing the neuromuscular block and discontinuing the inhaled anesthetic. The fentanyl infusion was continued as before. Once the patient had moved his or her feet to command, anesthesia was reestablished by inhaling a large concentration of the initial volatile anesthetic, titrated rapidly back to a target BIS of 60, and reestablishing neuromuscular blockade. This anesthetic was continued throughout the remaining SSEP-monitored surgery, targeting a BIS of 60, and was then used for the first two study SSEP measurements. After the first two study SSEP measurements, the anesthetic was switched from isoflurane to desflurane or from desflurane to isoflurane, as determined by the randomization process. The second anesthetic was continued until the end of the study, during which time two or three further SSEP measurements were made. Ten minutes was allowed to elapse after switching anesthetics and before preparing to begin the first SSEP collection with the patient breathing the second anesthetic. Patients continued to receive the same rate of infusion of fentanyl throughout the surgery, wake-up test, and study period. Esmolol and sodium nitroprusside were used as needed to produce controlled hypotension for blood loss control.
SSEPs were collected by using a Viking IV machine (Nicolet Biomedical, Madison, WI). Alternating stimulation of the left and right posterior tibial nerves was performed with subdermal needle electrodes. The rate of stimulation was 4.9 Hz. Intensity was initially set at 20 mA with a pulse width of 200 ms. If initial cortical responses appeared suboptimal, then the stimulus intensity was increased to a maximum of 30 mA. The higher stimulus potential was used if it resulted in a larger amplitude of the cortical responses. Cortical (P37-N45), subcortical (P31-N34), and popliteal fossa (Pf) potentials were recorded. Surface electrodes were placed according to the 10-20 system at Fpz′, Cz′, C3′, and C4′. An electrode was also placed near the cervical II level and served as a reference electrode in recording the subcortical potentials. Subdermal needle electrodes (Pf and Pf′) were used for recording the right and left Pf potentials. The cortical potentials were recorded by using the Fpz′-Cz′ channel. The alternate cortical channels Fpz′-C3′ and Fpz′-C4′ were used only if the initial cortical potentials recorded from the Fpz′-Cz′ channel were suboptimal. The subcortical potentials were recorded by using the cervical II-Fpz′ channel. The system bandpass was 30–3000 Hz. Each SSEP collection was averaged over 600–1000 repetitions (2).
After the induction of anesthesia, subjects then underwent their surgery, during which time maintenance anesthesia was provided with a constant fentanyl infusion and the initial inhaled anesthetic. The SSEP was monitored throughout this period for the purpose of patient care. The BIS was kept close to 60 (or slightly below) during this time. When the surgeon determined that SSEP monitoring was no longer needed, i.e., after the completion of spinal manipulation, the study period began.
The BIS was adjusted to a stable value of 60 ± 2, an SSEP collection was made, and the BIS was rechecked. Each SSEP collection was repeated once for the first anesthetic and was repeated at least once for the second anesthetic. The cortical and subcortical evoked potential latencies and amplitudes were recorded for both the left and right posterior tibial nerve-generated responses. If the patient’s BIS value changed by more than 10% during SSEP acquisition, then the SSEP data were discarded, and the collection was repeated.
The inhaled anesthetic concentrations were analyzed with either the Ohmeda Rascal 2 or the Datex Capnomac monitor. Age-specific minimum alveolar concentration (MAC) was calculated according to the method described by Mapleson (11). In addition, arterial blood pressure, end-tidal carbon dioxide partial pressure (ETco2), peripheral hemoglobin-oxygen saturation (Spo2), and esophageal temperature were recorded at measurement points.
A prospective power analysis was performed by using previous SSEP recordings from our department of neurophysiology to determine the underlying variability of the cortical SSEP. On this basis, our study was estimated to have power of 90% (P = 0.05) to detect a change in the cortical evoked potential amplitude of 50% with 16 patients, each acting as his or her own control. Our analysis proved conservative, because these historical SSEP amplitudes typically decreased over time, resulting in a higher sd being calculated over the course of a 4-h surgical recording period than would occur over a 40-min study period toward the end of surgery. Data were analyzed on an intention-to-treat basis, and all valid SSEP data were included (as described above). The duplicate (or triplicate) SSEP measurements were averaged for each anesthetic. Scalp recordings generated at each sampling time by stimulation of either the right or the left posterior tibial nerve were statistically indistinguishable, and measurements were therefore averaged together. Continuously distributed, between-patient data were analyzed with an unpaired Student’s t-test. The sex ratio used Fisher’s exact test according to the formula of Altman (12). Within-patient continuous data were analyzed with a paired Student’s t-test. Analysis was performed using the appropriate functions in an Excel 5 (Microsoft, Redmond, WA) spreadsheet and with StatView V (Abacus/SAS, Cary, NC). Data are presented as mean ± sd unless otherwise stated. A “p” value of 0.05 was considered significant.
Eighteen patients were included in the study. There were no postoperative neurological complications. The group demographics of those who received desflurane first were similar to the demographics of the group that received isoflurane first (Table 1). Propofol was used for induction in 5 of the 8 subjects in the ISO first group and in 6 of the 10 subjects in the DES first group. Incomplete data were obtained in two patients, both in the DES first group, who had bilaterally very low evoked cortical amplitudes that, after stimulation of the left posterior tibial nerve, were too low to interpret. These two patients were repeated; their incomplete data were included in the analysis (exclusion of their data did not change our findings). Thus, 8 patients in the ISO first group and 10 patients in the DES first group were included in the analysis. Despite equivalent BIS values, patients breathing desflurane had a lower amplitude of their cortical SSEP (P37-N45) compared with isoflurane (0.53 ± 0.30 μV versus 1.3 ± 0.80 μV; P = 0.014). Table 2 summarizes the conditions under which data were collected while the patient breathed the first study anesthetic. Table 3 shows the BIS and end-tidal anesthetic at each SSEP sampling time while breathing the second anesthetic. The sequential SSEP measurements made while breathing the second anesthetic did not show any significant change over time.
The results of the crossover component of the study are shown in Table 4 and Figure 1, A and B, and confirm the direction of change: inhaled desflurane results in a lower-amplitude cortical SSEP than inhaled isoflurane (0.78 ± 0.54 μV versus 0.93 ± 0.67 μV, respectively; P = 0.005; 95% confidence interval of the difference, 0.06–0.24 μV). There were no differences between isoflurane and desflurane in regard to the evoked potential latency (measured as the time between the stimulus and the P37 response) or the subcortical amplitude (P31-N34).
Previous studies of anesthetics have clearly shown a dose-related depression of the amplitude and desynchronization of the cortical SSEP at clinical doses (3–7). What is less clear is whether doses of different inhaled anesthetics that produce the same level of sedation (measured in this study by BIS) have equivalent effects on the cortical SSEP. Using a measure of consciousness state such as BIS to guide dosage of inhaled anesthetics would facilitate answering this important question. The answer to this question is particularly relevant in this era when many clinicians are altering their practice on the basis of information provided by such consciousness monitoring (13–16). We found that, at equivalent BIS values, patients breathing isoflurane had a higher evoked cortical SSEP amplitude than patients breathing desflurane.
In addition to comparing our two anesthetics independently, we included a randomized crossover design to control for interpatient variability. Switching between the two inhaled anesthetics in the same patient confirmed that cortical amplitudes were greater while breathing isoflurane.
Studies that have quantified depth of anesthesia have used MAC (3–7) rather than BIS. In our study, it appeared that patients breathing isoflurane required a larger MAC value to keep their BIS at 60 than those breathing desflurane. Although this was not statistically significant, it suggests that, on a MAC basis, patients breathing isoflurane were more deeply anesthetized than those breathing desflurane. BIS (9) and MAC (17) measure different physiological end-points. This has been previously demonstrated in children when, while breathing isoflurane or halothane at 1 end-tidal MAC, BIS was 56 while breathing halothane, whereas BIS was only 36 while breathing isoflurane (18). Similarly, in pigs at equi-MAC concentrations, BIS was higher during sevoflurane anesthesia than during isoflurane anesthesia (19). Despite the appearance that the patients receiving isoflurane received a slightly larger MAC concentration, they demonstrated less depression of their cortical SSEP amplitude. Thus BIS, MAC, and SSEP all appear to change independently in response to these two anesthetics. Whereas MAC is mediated at the spinal cord level (17), BIS and SSEP involve brain function and could be postulated to be better correlated than a MAC/SSEP comparison; however, our data do not support this idea, but rather suggest that, compared with isoflurane, breathing desflurane results in a divergence of the association between MAC, BIS, and cortical SSEP. Such interdrug differences exist in other physiological systems among various inhaled anesthetics, such as their relative effect on vascular smooth muscle and cardiac contractility, so these neurophysiological observations should not be surprising. Another neurophysiological comparison—BIS and the 95% spectral edge frequency—has also been suggested to be poorly correlated between sevoflurane and isoflurane (20), and a poor prediction between BIS and MAC when breathing isoflurane has also been demonstrated (21).
Our results are the first to use BIS to control for the dose of anesthetic given. Early studies examined the dose-response effect of inhaled anesthetics (3–6) and determined little if any difference at equivalent MAC concentrations. We had expected a similar result with the use of BIS to compare our two inhaled anesthetics, although after our study began, anesthesia with propofol resulted in a greater cortical SSEP amplitude than that with sevoflurane (22).
Peterson et al. (3) demonstrated a decrease in evoked cortical amplitude with increasing concentrations of isoflurane, enflurane, and halothane, with no statistical difference among the anesthetics, although the authors suggested that halothane was the least detrimental on the basis of loss of measurable amplitude with the other two anesthetics at larger concentrations. In contrast, Pathak et al. (4,5) published two similar studies. They found no significant difference between enflurane and isoflurane but noted that amplitudes became unmeasurable with halothane at 1.0 MAC.
These previous studies were confounded by the inclusion of 60% nitrous oxide, which is known to depress SSEPs (23). Nitrous oxide may be even more depressant on a MAC basis than desflurane (6), isoflurane, or enflurane (24).
More recently, Vaughan et al. (7) compared sevoflurane and desflurane under well controlled conditions by using median nerve stimulation and found that desflurane depressed the thalamocortical amplitude more than sevoflurane at equal MAC doses. This study also demonstrated augmentation of the subcortical (P15-N20) evoked amplitude at clinically useful sub-MAC concentrations of both anesthetics.
In contrast to the findings of Vaughan et al. (7) and our own, Rehberg et al. (25) reported that amplitude reduction was greater with isoflurane than with sevoflurane or desflurane at equal MAC concentrations; the last two anesthetics were similar. Of note in this dose-response study was the observation that most of the depression of cortical amplitude occurred between the control recording and when breathing 0.7 MAC, and much less effect occurred as the concentration increased to more than 0.7 MAC. Thus, it is possible that even small residual amounts of desflurane remaining while our patients breathed isoflurane could still produce a significant desflurane effect. A study by Bernard et al. (26) using posterior tibial stimulation observed no difference between isoflurane and desflurane. The authors stated that between-anesthetic comparisons could not be made because of their methodology. Of note, the patients receiving desflurane, although they received a concentration of desflurane similar to that of our subjects, had cortical amplitudes almost twice ours. Confounding this comparison is that the Bernard et al. (26) study was not randomized, the treatment groups were of dissimilar ages (and were generally older than ours), propofol was given to all patients as a background anesthetic, the stimulation rate was 2.8 Hz, and intermittent nitrous oxide was given (a 5- to 10-minute nitrous oxide washout period was allowed before each SSEP measurement in 100% oxygen).
Our study had some limitations. In particular, when using the crossover design, there is uncertainty regarding the amount of elimination of the first anesthetic when the SSEP is measured during the administration of the second anesthetic. We did not find that our gas analyzers produced reliable mixed anesthetic data at small concentrations. Our gas analyzer will detect a second (lower-concentration) anesthetic when that anesthetic is at a concentration of more than 0.3% and is present as at least 15% of the total anesthetic concentration (Datex-Ohmeda Capnomac Ultima manual, Section 7.1). Because our study used relatively small concentrations of inhaled anesthetic and because there is a large difference in the clinically useful concentration between desflurane and isoflurane because of their widely differing MACs, our anesthetic analyzer showed only one anesthetic within two or three minutes after switching anesthetics. We did not start SSEP measurements until at least 10 minutes had elapsed after the change of anesthetic. Data from a subsequent (as-yet unpublished) study comparing propofol with desflurane by using essentially the same protocol demonstrate that after switching from desflurane to propofol, the ET desflurane concentration had decreased to 0.37% at the time of the first propofol SSEP collection and to 0.27% by the third collection (neither was more than 10% of the concentration breathed just before the desflurane was discontinued). We would expect the same to be true of this study regarding desflurane, although the isoflurane concentration, based on the known kinetics, would be somewhat larger. Elimination kinetics of inhaled anesthetics have not been reported after prolonged exposure. Studies have been performed to determine washin and then washout after approximately 30 minutes of exposure (27,28). Table 3 shows the measured anesthetic concentrations over time while the patient was breathing the second anesthetic. It can be seen that while breathing desflurane (and eliminating isoflurane) that the ET anesthetic appears to increase. This suggests that continuing isoflurane elimination is occurring and that this necessitates incremental increases in the desflurane concentration to maintain the BIS at 60. The increase in measured desflurane is accentuated because a small percentage loss of isoflurane from the reported desflurane concentrations represents a relatively large MAC/sedation dose that requires a substantial increase in desflurane to maintain the BIS at 60. In contrast, patients breathing isoflurane (while eliminating desflurane) showed a steady concentration, thus suggesting the possibility that significant desflurane elimination had ceased. The concentration of isoflurane is, however, notably greater than the isoflurane concentration needed to maintain a BIS of 60 when it is used as the first anesthetic. A likely explanation of this is that even when the expired desflurane concentration has decreased to an insignificant MAC, e.g., 0.3%, when that 0.3% is included in the isoflurane concentration reported by the monitor, it increases the apparent isoflurane concentration being breathed by a significant MAC amount. Fortunately, the actual anesthetic concentrations during the crossover phase of this study were not of primary importance to us, because we used the BIS to determine the effect of the anesthetic. We accepted that the presence of mixed anesthetics contributed to anesthesia, but we would argue that on the basis of the kinetics of these inhaled anesthetics, the predominant effect on BIS and SSEP was caused by the anesthetic’s being delivered at the time of SSEP measurement. This situation also reflects the limited potential for improvement of the evoked cortical amplitude should the anesthesiologist choose to change from desflurane to isoflurane because of an unsatisfactory SSEP recording. What is most important to this randomized crossover phase of the study is the direction of change observed when each patient acts as his or her own control, and this confirmed the findings in the first part of the study.
The difference observed when switching between anesthetics represents a 15% decrease in amplitude when isoflurane is replaced by desflurane. Interestingly, the magnitude of the difference was larger when the cortical evoked potential amplitude was compared between isoflurane and desflurane in patients who were receiving their first anesthetic (patients in this comparison are not acting as their own controls). In this setting, breathing desflurane was associated with a reduction in cortical amplitude of the SSEP of 59% compared with isoflurane (Table 2). This difference was probably due to the impure nature of the anesthetic being inhaled after the crossover. Thus, although the direction of change was clear, the magnitude (15% vs 59%) of the change was uncertain. Although desflurane was clearly associated with lower cortical evoked potential amplitudes than isoflurane, within-patient factors were also important. Patients who had low cortical evoked potential amplitudes while breathing one anesthetic also had low amplitudes while breathing the other. We never observed a patient who had low cortical evoked potential amplitudes while breathing desflurane but high amplitudes when breathing isoflurane.
In conclusion, our study compared the effects of isoflurane with desflurane on the SSEP during correction of idiopathic spinal scoliosis in teenaged pediatric patients. The amplitude of the cortical evoked potential was less while breathing desflurane than while breathing isoflurane, without apparent effects on subcortical evoked potential amplitude or cortical evoked potential (P37) latency.
We thank Drs. Richard C. Henderson and Edmund R. Campion (Department of Pediatric Orthopedic Surgery, University of North Carolina, Chapel Hill) for their support.
1. Blinder J, El-Mansouri M, Connelly NR. Survey of anesthesia practice in spine surgery patients in the United States. J Anesth 2003;17:8–12.
2. Guideline eleven: guidelines for intraoperative monitoring of sensory evoked potentials. J Clin Neurophysiol 1994;11:77–87.
3. Peterson DO, Drummond JC, Todd MM. Effects of halothane, enflurane, isoflurane, and nitrous oxide on somatosensory evoked potentials in humans. Anesthesiology 1986;65:35–40.
4. Pathak KS, Amaddio MD, Scoles PV, et al. The effects of halothane, enflurane, and isoflurane in nitrous oxide on multilevel somatosensory evoked potentials. Anesthesiology 1989;70:207–12.
5. Pathak KS, Ammadio M, Kalamchi A, et al. Effects of halothane, enflurane, and isoflurane on somatosensory evoked potentials during nitrous oxide anesthesia. Anesthesiology 1987;66:753–7.
6. Schindler E, Muller M, Zickmann B, et al. Modulation of somatosensory evoked potentials under various concentrations of desflurane with and without nitrous oxide. J Neurosurg Anesthesiol 1998;10:218–23.
7. Vaughan DJA, Thornton C, Wright DR, et al. Effects of different concentrations of sevoflurane and desflurane on subcortical somatosensory evoked responses in anaesthetized, non-stimulated patients. Br J Anaesth 2001;86:59–62.
8. Flaishon R, Windsor A, Sigl J, Sebel PS. Recovery of consciousness after thiopental or propofol: bispectral index and isolated forearm technique. Anesthesiology 1997;86:613–9.
9. Glass PS, Bloom M, Kearse L, et al. Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane, and alfentanil in healthy volunteers. Anesthesiology 1997;86:836–47.
10. Kearse L, Rosow C, Zaslavsky A, et al. Bispectral analysis of the electroencephalogram predicts conscious processing of information during propofol sedation and hypnosis. Anesthesiology 1998;88:25–34.
11. Mapleson WW. Effect of age on MAC in humans: a meta-analysis. Br J Anaesth 1996;76:179–85.
12. Altman DG. Practical statistics for medical research. Boca Raton, FL: CRC Press, 1999:253–7.
13. Chin KJ, Yeo SW. A BIS-guided study of sevoflurane requirements for adequate depth of anaesthesia in Caesarean section. Anaesthesia 2004;59:1064–8.
14. Kreuer S, Bruhn J, Larsen R, et al. Application of bispectral index and Narcotrend index to the measurement of the electroencephalographic effects of isoflurane with and without burst suppression. Anesthesiology 2004;101:847–54.
15. Liu SS. Effects of Bispectral Index monitoring on ambulatory anesthesia: a meta-analysis of randomized controlled trials and a cost analysis. Anesthesiology 2004;101:311–5.
16. White PF, Ma H, Tang J, et al. Does the use of electroencephalographic bispectral index or auditory evoked potential index monitoring facilitate recovery after desflurane anesthesia in the ambulatory setting? Anesthesiology 2004;100:811–7.
17. Rampil IJ, Mason P, Singh H. Anesthetic potency (MAC) is independent of forebrain structures in the rat. Anesthesiology 1993;78:707–12.
18. Davidson AJ, Czarnecki C. The Bispectral Index in children: comparing isoflurane and halothane. Br J Anaesth 2004;92:14–7.
19. Martin-Cancho MF, Lima JR, Luis L, et al. Bispectral index, spectral edge frequency 95%, and median frequency recorded for various concentrations of isoflurane and sevoflurane in pigs. Am J Vet Res 2003;64:866–73.
20. Kurehara K, Horiuchi T, Takahashi M, et al. Relationship between minimum alveolar concentration and electroencephalographic bispectral index as well as spectral edge frequency 95 during isoflurane/epidural or sevoflurane/epidural anesthesia. Masui 2001;50:512–5.
21. Kurehara K, Takahashi M, Kitaguchi K, Furuya H. The relationship between end-tidal isoflurane concentration and electroencephalographic bispectral index during isoflurane/epidural anesthesia. Masui 2002;51:642–7.
22. Boisseau N, Madany M, Staccini P, et al. Comparison of the effects of sevoflurane and propofol on cortical somatosensory evoked potentials. Br J Anaesth 2002;88:785–9.
23. Sloan TB, Koht A. Depression of cortical somatosensory evoked potentials by nitrous oxide. Br J Anaesth 1985;57:849–52.
24. McPherson RW, Mahla M, Johnson R, Traystman RJ. Effects of enflurane, isoflurane, and nitrous oxide on somatosensory evoked potentials during fentanyl anesthesia. Anesthesiology 1985;62:626–33.
25. Rehberg B, Ruschner R, Fischer M, et al. Concentration-dependent changes in the latency and amplitude of somatosensory-evoked potentials by desflurane, isoflurane and sevoflurane. Anasthesiol Intensivmed Notfallmed Schmerzther 1998;33:425–9.
26. Bernard JM, Pereon Y, Fayet G, Guiheneuc P. Effects of isoflurane and desflurane on neurogenic motor- and somatosensory-evoked potential monitoring for scoliosis surgery. Anesthesiology 1996;85:1013–9.
27. Yasuda N, Lockhart SH, Eger EI II, et al. Kinetics of desflurane, isoflurane and halothane in humans. Anesthesiology 1991;74:489–98.
© 2005 International Anesthesia Research Society
28. Yasuda N, Lockhart SH, Eger EI II, et al. Comparison of kinetics of sevoflurane and isoflurane in humans. Anesth Analg 1991;72:316–24.