sufentanil anaesthesia is related to changes in the midlatency auditory evoked response

British Journal of Anaesthesia 89 (3): 376-81 (2002)

BJA

CLINICAL INVESTIGATIONS

Impaired explicit memory after recovery from propofol/sufentanil anaesthesia is related to changes in the midlatency auditory evoked response I. Rundshagen1*, K. Schnabel2 and J. Schulte am Esch3

^Corresponding author Background. Midlatency auditory evoked responses (MLAER) can distinguish different stages of anaesthesia. We studied MLAER during emergence from propofol/sufentanil anaesthesia in relation to recovery of explicit memory. Methods. MLAER were recorded in 29 healthy patients before and during anaesthesia -and during emergence until the patients opened their eyes spontaneously. After a structured interview the next day, patients were classified into those with and without explicit memory of the recovery period. Latencies Na, Pa and Nb and the peak-to-peak amplitudes NaPa and PaNb were compared between the groups by multivariate analysis of variance. Results are mean (SD). Results. At eye opening (37 (12) min after the end of anaesthesia) the latency Nb (47 (5) compared with 41 (5) ms; P<0.00l) was prolonged and the amplitude PaNb (1.3 (0.8) compared with I (0.5) ms; P=0.012) was greater than the baseline value, respectively. The Nb latency was significantly shorter in patients with explicit memory (49 (2) ms compared with 45 (l);P=0.04l). Conclusions. Large intra- and inter-individual variability in the MLAER values limited their ability to predict memory responses in individual patients during emergence from propofol/ sufentanil anaesthesia. BrJ Anaesth 2002; 89: 376-81 Keywords: anaesthesia i.v., propofol; analgesics opioid, sufentanil; brain, evoked potentials; memory Accepted for publication: April 24, 2002

Measurement of anaesthetic effects on the brain can assist more precise dosage and reduce side-effects. Values measured from the spontaneous electroencephalogram (EEG), such as the bispectral index, or from the evoked EEG, such as auditory evoked responses, may indicate anaesthetic effects during anaesthesia.1 Auditory evoked responses (AER) are derived by averaging the EEG after repeated short auditory stimuli, which gives a highly reproducible sequence of waveforms. The brain stem auditory evoked responses (BAER) change little with anaesthetics.2 The waveforms following the BAER components, the midlatency auditory evoked responses

(MLAER), are more sensitive to anaesthetics, with characteristic changes in latencies and amplitudes.3 Several studies suggest that the MLAER can distinguish different states of consciousness. Davies and colleagues found consistent changes in MLAER latencies during repeated transitions from unconsciousness to consciousness during sedation with propofol in combination with spinal anaesthesia.4 A threshold value of the MLAER latency Nb of 53 ms discriminated between the presence and the absence of an eyelash reflex with a sensitivity of 100% and a specificity of 96% during propofol sedation before surgery.5 MLAER waveforms may indicate the ability to form

© The Board of Management and Trustees of the British Journal of Anaesthesia 2002

Downloaded from http://bja.oxfordjournals.org/ at Australian National University on March 16, 2015

department of Anaesthesiology, University Hospital Charite, Campus Charite Mitte, Schumannstrasse 20/21, D-10117 Berlin, Germany. Department of Psychology, University of Michigan, Ann Arbor, USA. Department of Anaesthesiology, University Hospital Eppendorf, Hamburg, Germany

MLAER and memory during recovery

implicit memories.6 An increase of the Pa latency of <12 ms distinguished patients with and without implicit memory after operation, with a sensitivity of 100% and a specificity of 77%. We found that explicit memory performance in early recovery could be distinguished in patients waking up from isoflurane/nitrous oxide anaesthesia, using changes in the median nerve somatosensory evoked response (MnSSER) to indicate recovery of explicit memory.7 The present study was designed to measure MLAER and explicit memory during recovery from propofol/sufentanil anaesthesia. We wished to compare the MLAER waveforms between patients with and without explicit memory for this time.

Patients We studied 29 patients (18 males, 11 females; mean age 40 (18-61) yr, height 1.76 (0.10) m, weight 76 (12) kg; ASA I—II), scheduled for elective eye surgery lasting ~1 h, after approval from the local ethics committee and with written informed consent. We excluded patients with neurological diseases or hearing abnormalities, patients receiving centrally acting drugs, and those who had an ASA grade >II.

Anaesthesia and recovery The patients received midazolam 7.5 mg p.o. for premedication 45 min before anaesthesia. After induction of anaesthesia with propofol 2 mg kg"1, sufentanil 0.4 |ig kg"1 and vecuronium 0.1 mg kg"1, the trachea was intubated and anaesthesia maintained with propofol 8 mg kg"1 h"1. The inspired oxygen concentration was 0.3. Additional doses of sufentanil 0.2 (ig kg"1 were given when needed. At the end of surgery and after recording the MLAER, the propofol infusion was stopped. One hundred per cent oxygen was given using a fresh gas flow at 3 litre min"1 and the patients were extubated when breathing was adequate. After the extubation the patients received oxygen via a face mask.

Auditory evoked responses MLAER recordings were made in a standard manner with an Evomatic 4000® system (Dantec, Copenhagen, Denmark). After defining the individual hearing threshold, a stimulus intensity of -70 dB above this threshold was chosen and kept constant throughout the whole study. A random click to both ears via headphones was used. The stimulus frequency was 8 Hz. The MLAER waveforms were recorded on two amplifier channels using cup electrodes (zinc/lead) placed over Cz (international 10-20 system) and both mastoid bones. Impedances were kept below 5 k£2. A bandpass of 0.02-2 kHz was used and 2000 stimuli were averaged for each response and stored on disk for later

Measurements The patients were shown the procedure of MLAER recording on the day before surgery, and baseline values were obtained (Awake). Duplicate baseline recordings were performed. After surgery had finished, MLAER were recorded during steady-state propofol anaesthesia (Anaesth), and every 5-10 min during emergence from anaesthesia. Since the duration of recovery from anaesthesia varied from patient to patient, clinically comparable stages were defined for comparison of the MLAER components. Pre-EXT was defined as the last MLAER recording before extubation, Post-EXT as the first recording after extubation when the patients opened their eyes on command, and Recovery when the patients opened their eyes spontaneously and could state the name of an object shown to them.

Clinical measurements We noted heart rate (beats min"1), non-invasive mean arterial pressure (mm Hg), percutaneous oxygen saturation (%) and body temperature when MLAER were recorded. End tidal carbon dioxide partial pressure (PE'CO2) w a s measured during and after anaesthesia, and we noted the duration of anaesthesia, the time of extubation and the time of recovery.

Assessment of memory When the patients had regained consciousness and opened their eyes spontaneously, they were asked to name precisely an object shown to them. A red booklet was opened and closed in front of them and they were asked to keep this booklet in mind. On the day after surgery a structured interview was used to assess the patients' memory of the immediate recovery period. A broad variety of questions were posed. The relevant questions are listed in Table 1. According to their responses the patients were classified into a group without recall (No-MEM) and a group with explicit recall for the immediate recovery period (MEM). Table 1 Structured interview 24 h after anaesthesia What was the first thing you remembered after waking up from anaesthesia? Do you remember anything being shown to you when you woke up? Do you remember whether you heard clicks? Were you asked to remember anything during the time that you recovered?

377

Downloaded from http://bja.oxfordjournals.org/ at Australian National University on March 16, 2015

Methods

analysis. MLAER were analysed for 90 ms after the stimulus. During anaesthesia and recovery, a period of 180 ms post-stimulus was measured to detect excessive prolongation of the waveforms. The peak latencies of the brainstem component, V, of the two negative MLAER components, Na and Nb, and the positive peak Pa between were measured with a software package (EvoPC®; Miiller, Hamburg, Germany). Peak-to-peak amplitudes NaPa and PaNb were measured.

Rundshagen et al.

Statistics

Results Clinical findings and patient characteristics Values are given as mean (SD). The surgical procedure lasted for 57 (26) min, and the total time of anaesthesia was 97 (28) min. Extubation was done 21 (9) min after the propofol infusion was stopped. The patients were able to identify the object shown to them 37 (13) min after the propofol infusion was stopped. Clinical measurements are listed in Table 2. The heart rate decreased during Anaesth and Pre-EXT compared with Awake, and the mean arterial blood pressure (MAP) decreased during Anaesth. Oxygen saturation, PE'CO2 a n d body temperature showed minor changes. Two patients were not interviewed the day after surgery because they had been discharged. They were excluded from further analysis. Thirteen patients had explicit memory of the immediate recovery period, either recalling the shown object or the click stimulation of AER recording. Three of these patients recalled extubation. Fourteen patients did not remember anything of the study procedure. There was no difference in the characteristics or clinical measurements

MLAER parameter during emergence from anaesthesia The brainstem component V, which serves as a reference for an artefact-free recording, was identified in all patients when awake. The postauricular reflex was seen in some but not all of the patients, and the amplitudes could be seen to differ markedly. After this there were two negative and one positive waves of the MLAER in all patients except in two: one did not show the component Nb, and the other did not show the components Pa and Nb. The assumption of a normal distribution for the MLAER latencies was supported by the Kolmogorov-Smirnov test. The amplitude NaPa was not normally distributed, so the amplitudes were analysed by non-parametric tests (the Friedman and Wilcoxon tests). Multivariate analysis showed a significant difference for the latency components comparing the different measurements CP
8070^ 60H g>

M 50 0

i

40

LU


Anaesth

Pre-EXT Post-EXT

Recovery

Fig 1 Means (SD) of MLAER latency components Na, Pa and Nb during the different measurements: awake, during propofol/sufentanil anaesthesia, before extubation, after extubation and at recovery. *P<0.05 vs awake.

Table 2 Clinical measurements. Means (SD) of measurements when awake, during propofol anaesthesia, before extubation, after extubation and at recovery. MAP=mean arterial pressure, 5o2=pulse oximeter reading, PE'co,=end tidal carbon dioxide partial pressure. Temperature was measured rectally

Heart rate (beats min ) MAP (mm Hg) So2 (%) PE'CO, (mm

Awake

Anaesth

Pre-EXT

Post-EXT

Recovery

68 (13) 93 (11) 98(1)

62(11) 86 (12) 99(1) 35 (3) 36.4 (0.3)

59(9) 90(11) 99(1) 36(3) 36.4 (0.3)

69 (12) 100 (14) 99(1) 43 (6) 36.3 (0.3)

68 (11) 96 (12) 99(1) 41.2(5) 36.4 (0.4)

Hg)

Temperature (°C)

36.5 (0.4)

378

Downloaded from http://bja.oxfordjournals.org/ at Australian National University on March 16, 2015

Latencies and peak-to-peak amplitudes NaPa and PaNb were calculated. The distribution of data was tested with the Kolmogorov-Smirnov test. The changes in the MLAER data were analysed using multivariate analysis of variance (ANOVA; Hotellings T-square; repeated measurement design) if the data were normally distributed. If the distribution was not normal, the data were analysed by non-parametric tests (Friedman test). All five measurements were included in a multivariate analysis giving a full two-factor-within design (three latencies X five time points). For descriptive purposes, a posteriori paired Mests were used to compare MLAER components. We tested for correlation between the MLAER components, and the clinical measurements. Inter-group comparisons for the MLAER components were made using multivariate ANOVA for repeated measurements (Hotelling's T-square) using Awake, Anaesth and Recovery values. Univariate comparisons were also made for the MLAER components at Awake, Anaesth and Recovery. The Wilcoxon test was used if a normal distribution was not present. .P<0.05 was taken to represent significance for all statistical tests.

between these two memory groups. The dose of sufentanil, the duration of anaesthesia, the time to extubation and the duration of recovery showed no differences.

MLAER and memory during recovery Table 3 MLAER latencies and amplitudes. Data are means (SD) at awake, during propofol anaesthesia, before extubation, after extubation and at recovery. *P<0.05 vs awake

Latency Na (ras) Latency Pa (ms) Latency Nb (ms) Amplitude NaPa (uV) Amplitude PaNb (LJV)

Awake

Anaesth

Pre-EXT

Post-EXT

Recovery

17.2 (0.3) 30.1 (0.6) 41.4(1.0) 3.0 (0.6) 1.0(0.1)

21.3 (0.6)* 40.4 (0.9)* 65.9 (1.9)* 0.5 (0.1)* 0.4(0.1)*

18.8 (0.3)* 32.4 (0.6)* 53.7 (1.3)* 1.3 (0.3)* 1.0 ( 0.1)

17.7 (0.4) 31.4 (0.5) 48.6 (1.0)* 2.4 (0.6) 1.4 (0.1)

17.7 (0.3) 31.3 (0.6) 47.3 (1.0)* 3.1 (0.6) 1.3 (0.2)*

times (P<0.001). The absolute values for the peak-to-peak amplitudes are listed in Table 3. The brainstem component V was seen during Anaesth with propofol in all patients. Compared with the awake measurements, the MLAER latencies were prolonged (P<0.001) and the amplitudes diminished (P<0.001). In three patients the component Nb was completely suppressed, and Pa was suppressed in one of these. These changes gradually reversed during recovery. Measurements were made before extubation in 26 patients. The latencies remained prolonged compared with awake (P<0.05), but decreased significantly compared with anaesthesia (P<0.001). The amplitudes of NaPa and PaNb increased significantly in comparison to anaesthesia, but NaPa remained reduced compared with awake (F<0.001). After extubation MLAER were recorded in 20 patients, when the patients became responsive and opened their eyes at command. The latencies Na and Pa and the amplitudes regained baseline levels, but the latency Nb remained prolonged in comparison to before anaesthesia (P=0.001). When the patients opened their eyes spontaneously, MLAER components were seen in all patients again, except Pa, and Pa and Nb in the patients who did not have these components before anaesthesia. The latency Nb remained prolonged in comparison to preoperative measurement (P<0.001). The amplitude PaNb exceeded the baseline value (P<0.05).

(i)

(iii)

I

I

-10

0

90 ms

(i)

(•i)

(iii)

MLAER measurements and memory There was no difference between MLAER values before surgery between patients who later had explicit recall (n=13) and those who did not (n=14). The time of anaesthesia, time to extubation, duration of recovery and the dose of sufentanil were not different between the groups. At Recovery, in contrast, the latency of Nb was significantly less in the patients with explicit memory compared with the patients without (49.3 (1.8) ms vs 44.9 (1.3); P=0.041). The other MLAER components did not differ. Figure 2 shows two original AER traces at Awake and Recovery: one from a patient without and one with explicit memory. The individual Nb latencies at Awake and Recovery for the two memory groups are shown in Figure 3. Eleven data points only are shown for the group MEM, because two

-10

0

90 ms

Fig 2 AER tracings of two patients. The brain stem component V and the three midlatency components are indicated. The upper traces (A) show the AER of a female patient who did not remember recovery. The lower traces (B) show the AER of a male patient who had explicit memory. The different measurements are: (i) awake, (ii) during propofol/sufentanil anaesthesia after surgery, and (iii) when the patients opened their eyes spontaneously. Vertical lines indicate the Nb latencies. EEG-Montage: Cz vs linked mastoid, binaural click stimulation.

patients in that group did not have an Nb component during Awake and Recovery.

379

Downloaded from http://bja.oxfordjournals.org/ at Australian National University on March 16, 2015

(ii)

Rundshagen et al.

65-

No-MEM

MEM

555045403530 Awake

Recovery

Awake

Recovery

Discussion We found graded changes of MLAER during emergence from anaesthesia. The anaesthetic-induced changes in the MLAER recovered when the patients regained consciousness and were able to identify an object correctly. The persistent increase in the latency Nb indicated impaired processing of auditory signals 38 min after anaesthesia with propofol/sufentanil. The latency of Nb was shorter in the patients who had explicit recall compared with patients without explicit memory. These findings support previous reports that BAER and MLAER show a decrease in the amplitudes and an increase in latencies during anaesthesia. 289 BAER and MLAER were measured during increments of anaesthesia with propofol.10 In contrast to the BAER, the MLAER were affected markedly by propofol (7 mg kg"1 IT1), and latencies and peaks could not be measured. This is in contrast to the present study, where Nb was completely suppressed in three patients and Pa was suppressed in one patient only. Depression of MLAER is in part reversed during surgical stimulation, which could explain these different results.11 12 Tooley and colleagues used MLAER changes to establish a dose response curve using propofol as a sole agent in patients without surgical stimulation.5 A cut-off value of 53 ms for the Nb latency had a sensitivity of 100% and a specificity of 96% as a discriminator of eyelash response vs no response. Concentrations >6 \ig ml"1 attenuated the auditory response to an extent that made it difficult to determine the components. Care was needed to obtain AER in awake patients without premedication, because muscle artefact affected the signal. Coherent averaging gave the most reliable estimate of the AER. In the present study we recorded MLAER the day before surgery to familiarize the patients with the procedure, and repetitive stimulation reduced the effect of muscle artefact. We could not obtain MLAER during and immediately after extubation because of muscle artefact. During recovery the averaging time was sometimes prolonged because artefacts were present.

380

Downloaded from http://bja.oxfordjournals.org/ at Australian National University on March 16, 2015

Fig 3 The individual Nb latencies are indicated at awake and at recovery, when the patients opened their eyes after anaesthesia. The patients of the group No-MEM did not remember recovery, while the patients of the group MEM had explicit recall.

Coherent averaging might have improved the extraction of MLAER during recovery. The mean latency of Nb was -54 ms, when the patients were tolerating the tracheal tube immediately before extubation. It decreased to -48 ms after extubation, when the patients opened their eyes at command, which supports the observations of Tooley and colleagues.5 Our study does not indicate a clear threshold because of the large inter-individual overlap in the Nb data. Muscle artefacts during recovery could affect visual interpretation of the components. However, large inter-individual differences in the MLAER components have been described in the awake state, and MLAER are rarely used to diagnose hearing problems.8 13 In a previous study we recorded median nerve somatosensory-evoked responses during emergence from anaesthesia.7 When the patients identified an object shown to them when waking from anaesthesia, the amplitude P25N35 exceeded the awake value. In the present study the amplitude PaNb was greater than the awake values. These changes may reflect an effect of cortical arousal by surgery, but several factors may contribute to the results. The patients had received midazolam for premedication and sufentanil intraoperatively. In contrast to the increase in amplitudes, a persistent delay in the latency Nb was found to be present when the patients had regained consciousness, which could be related to explicit memory performance. In previous studies with median nerve somatosensory-evoked responses we found that midlatency MnSSER components were prolonged and correlated with explicit memory. 714 The cortical components of evoked responses, >30 ms post-stimulus, are more sensitive to residual anaesthetics than the earlier components and need more time to recover, irrespective of the stimulation mode. These indicate impairment of more complex cognitive processing such as implicit or explicit memory processing. The coherent frequency of the AER could be related to categorical memory tests during isoflurane administration.15 During sedation with propofol, memory performance correlates more strongly with changes in AER amplitude P3 than with the drug concentration.16 However, AER components may not reliably indicate memory during different states of consciousness.17 Irrespective of the significant group effects, our results suggest a low predictive value because the inter-individual variation of Nb latency is considerable. Computer-assisted measurements of the AER, such as the AER index or use of moving averaging techniques, have been developed.1819 During recovery from an induction dose of propofol, concentration-related effects on the AER index were found, but a threshold was found in seven of 22 patients.20 The AER index returned to awake levels only when the patients had already developed clinical signs of arousal. In 10 female patients anaesthetized with propofol and nitrous oxide, there was no difference between the awake AER index and the values when the patients opened their eyes at command.21 Further studies are needed to show

MLAER and memory during recovery

if the predictive value of the MLAER is improved by computer-assisted techniques. As Thornton states: "A brain signal recorded at one point in time will not necessarily predict future behaviour accurately if events or treatments which occur in the period between the two time points can influence behaviour".3

I I Thornton C, Konieczko K, Jones JG, et al. Effect of surgical stimulation on the auditory evoked response. BrJ Anaesth 1988; 60: 372-8 12 Crabb I, Thornton C, Konieczko KM, et al. Remifentanil reduces auditory and somatosensory evoked responses during isoflurane anaesthesia in a dose-dependent manner. Br J Anaesth 1996; 76: 795-801 13 McPherson D, Starr E. Auditory evoked potentials in the clinic. In: Halliday A M , ed. Evoked Potentials in Clinical Testing. Edinburgh:

References

381

14

15

16

17

18

19

20

21

Downloaded from http://bja.oxfordjournals.org/ at Australian National University on March 16, 2015

1 Drummond JC. Monitoring depth of anesthesia: with emphasis on the application of the bispectral index and the middle latency auditory evoked response to the prevention of recall. Anesthesiology 2000; 93: 876-82 2 Savoia G, Esposito C, Belfiore F, Amantea B, Cuocolo R. Propofol infusion and auditory evoked potentials. Anaesthesia 1988; 43 (Suppl): 46-9 3 Thornton C, Sharpe RM. Evoked responses in anaesthesia. BrJ Anaesth 1998; 81:771-81 4 Davies FW, Mantzaridis H, Kenny GN, Fisher AC. Middle latency auditory evoked potentials during repeated transitions from consciousness to unconsciousness. Anaesthesia 1996; 51:107-13 5 Tooley MA, Greenslade GL, Prys-Roberts C. Concentrationrelated effects of propofol on the auditory evoked response. BrJ Anaesth 1996; 77: 720-6 6 Schwender D, Kaiser A, Klasing S, Peter K, Poppel E. Midlatency auditory evoked potentials and explicit and implicit memory in patients undergoing cardiac surgery. Anesthesiology 1994; 80: 493-501 7 Rundshagen I, Schnabel K, Schulte am Esch J. Median nerve evoked responses and explicit memory during recovery from isoflurane/nitrous oxide anesthesia. Can J Anaesth 2000; 47: 494-502 8 Ozdamar O, Kraus N. Auditory middle-latency responses in humans. Audiology 1983; 22: 34—49 9 Thornton C, Konieczko KM, Knight AB, et al. Effect of propofol on the auditory evoked response and oesophageal contractility. BrJ Anaesth 1989; 63: 41 1-7 10 Chassard D, Joubaud A, Colson A. et al. Auditory evoked potentials during propofol anaesthesia in man. BrJ Anaesth 1989; 62: 522-6

Churchill Livingstone, 1993; 359-81 Rundshagen I, Schnabel K, Schulte am Esch J. Midlatency median nerve evoked responses during recovery from propofol/ sufentanil total intravenous anaesthesia. Acta Anaesthesiol Scand 2000; 44: 313-20 Munglani R, Andrade J, Sapsford DJ, Baddeley A, Jones JG. A measure of consciousness and memory during isoflurane administration: the coherent frequency. BrJ Anaesth 1993; 71: 633^1 Reinsel RA, Veselis RA, Wronski M, Marino P. The P300 eventrelated potential during propofol sedation: a possible marker for amnesia? Br J Anaesth 1995; 74: 674-80 Thornton C, Sharpe RM. The auditory evoked responses and memory during anesthesia. In: Ghoneim MM, ed. Awareness during Anesthesia. Oxford: Butterworth Heinemann, 2001; 117-28 Jensen EW, Nebot A, Caminal P, Henneberg SW. Identification of causal relations between haemodynamic variables, auditory evoked potentials and isoflurane by means of fuzzy logic. Br J Anaesth 1999; 82: 25-32 Mantzaridis H, Kenny GN. Auditory evoked potential index: a quantitative measure of changes in auditory evoked potentials during general anaesthesia. Anaesthesia 1997; 52: 1030-6 White M, Schenkels MJ, Engbers FH, et al. Effect-site modelling of propofol using auditory evoked potentials. BrJ Anaesth 1999; 82: 333-9 Doi M, Gajraj RJ, Mantzaridis H, Kenny GN. Relationship between calculated blood concentration of propofol and electrophysiological variables during emergence from anaesthesia: comparison of bispectral index, spectral edge frequency, median frequency and auditory evoked potential index. BrJ Anaesth 1997; 78: 180-4