- Email: [email protected]
fluanisone−midazolam anaesthesia in rats
Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
R E S E A R C H PA P E R
Evaluation of auditory evoked potentials to predict depth of anaesthesia during fentanyl/fluanisonemidazolam anaesthesia in rats LM Antunes
MSc, DVM,
JV Roughan
BSc, PhD
& PA Flecknell
MA, Vet MB, PhD, DLAS, Dip ECVA
Comparative Biology Centre, Medical School, University of Newcastle upon Tyne, UK
Correspondence: Luis Antunes, Comparative Biology Centre, Medical School, University of Newcastle upon Tyne NE2 4HH, UK.
Abstract Objectives To assess a method for monitoring depth of anaesthesia using components of middle latency auditory evoked potential (AEP) waveforms during anaesthesiawith fentanyl/£uanisoneand midazolam. Study design Prospective observational study. Animals Five female Wistar rats weighing between 210 and 250 g. Methods Implanted electrodes were used to record AEPs in animals receiving ¢ve doses of anaesthetic. Recordings were made at 5 minutes post-injection (deep anaesthesia; no pedal withdrawal response, PWR) and then at 25 minutes (light anaesthesia; strong PWR). Responses showed ¢ve characteristic peaks occurring at 11, 14, 23, 42 and 68 ms that were measured for latency of occurrence and peak amplitude. Results Auditory evoked potential peaks P14, N23 and P42 were increased signi¢cantly in latency with successive anaesthetic injections [avg. F(1,4) ¼ 12.53, p < 0.001; avg. F(1,4) ¼ 10.6, p < 0.001; avg. F(1,4) ¼ 3.9, p ¼ 0.02, respectively]. Peak N23 showed a signi¢cant reduction in latency during the 20 minute recovery period following both the ¢rst and second anaesthetic injections (t(3) ¼ 7.52, p ¼ 0.005; t(4) ¼ 5.17, p ¼ 0.007, respectively). Peak P42 occurred signi¢cantly earlier 20 minutes following the second anaesthetic injection (t(4) ¼ 4.75, p ¼ 0.009). The mean overall depth of anaesthesia assessed using PWR scores was signi¢cantly correlated with the mean latency of peak N23, such that as the strength of PWR increased, N23 occurred 196
signi¢cantly earlier (r ¼ 0.99, p ¼ 0.01). The amplitude di¡erence between peaks N23 and P42 increased after the second and third drug administrations [avg. F(1,4) ¼ 10.65, p ¼ 0.031 and avg. F(1,4) ¼ 11.24, p ¼ 0.028, respectively]. Conclusion The characteristics of these peaks, and in particular latency of peak N23, may provide a useful tool for assessing depth of anaesthesia produced by this, and possibly other anaesthetic agents. Keywords anaesthesia, auditory evoked potential, fentanyl/ £uanisone/ midazolam, rat.
Introduction The Animals (Scienti¢c Procedures) Act (1986) and European Directive (86/609/EEC) de¢ne the care requirements of experimental animals undergoing surgical procedures and impose a greater obligation on researchers to ensure the provision of adequate anaesthesia. Several recent studies in humans have considered changes in auditory evoked potential (AEP) activity as a potentially useful tool for evaluating anaesthetic depth (Thornton 1991). Auditory evoked responses are of short (approximately <100 millisecond) duration, and are generated as an average of EEG activity following repetitive stimulus presentation. Their amplitude and latency are thought to re£ect the activation state of neuronal populations (Fox & O’Brian 1965). Evoked responses are considered powerful diagnostic tools as they are frequently closely correlated with behavioural, cognitive and clinical states (Bullock 1981). Auditory evoked responses are enhanced in amplitude and
Depth of anaesthesia monitoring with AEP LM Antunes et al.
reduced in latency when general arousal levels are high, contrary to the circumstances encountered during anaesthesia. Accordingly, dose-dependent amplitude depression and increased latency of AEP components during anaesthesia are now generally accepted indices of depth for some anaesthetics (e.g. volatile anaesthetics and propofol) (Thornton et al. 1989; Schwender et al. 1995, 1997). Relatively few animal studies have been conducted, and of these, most have assessed techniques such as respiratory rate, loss of the righting re£ex and strength of pedal withdrawal response (PWR) (Whelan & Flecknell 1992). This may be due, in part, to reports of poor correlation between EEG parameters and the strength of these somatic responses, for example during pentobarbital anesthesia (Haberham et al.1999) and iso£urane anaesthesia in rats (Rampil & Laster1992). Because of the increasing use of anaesthetic techniques employing neuromuscular blocking agents in which these types of assessment are unavailable, there is now a requirement for a more reliable and objective assessment technique. Previous studies used a broad spectrum of anaesthetic agents across a range of species. The con£icting data produced may have resulted from the lack of a su⁄ciently systematic approach to anaesthesia depth assessment; thus, AEP measurements may represent a useful alternative. Here we describe the results of a pilot study to characterize the average EEG response of rats to a series of click stimuli, delivered during anaesthesia induced with iso£urane and subsequently maintained with increasing doses of fentanyl/£uanisone^midazolam, a neuroleptanalgesic combination commonly used to anaesthetize laboratory rats (Flecknell & Mitchell1984).
Materials and methods All procedures were authorized under the Animals (Scienti¢c Procedures) Act (1986) and were approved by the institutional ethics committee. Five female outbred Wistar rats (Charles River, UK) weighing between 210 and 250 g and free from recognized respiratory pathogens were used. Animals were grouphoused in solid-£oored caging (RC1, North Kent Plastics Ltd., Erith, UK;56 cm 38 cm, height18 cm). Sawdust (‘Gold Chip’, BS and S Ltd., Edinburgh, UK) was used as bedding. Stocking densities were within UK-speci¢ed guidelines [Animals (Scienti¢c Procedures) Act 1986; 1989]. Room temperature was controlled at 20 2 8C. A 9:15 hours light/dark cycle was maintained (lights on 08:00 hours o¡ 17 : 00 hours). Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
Animals received a commercial pellet diet (R and M no. 1, Special Diets Services Ltd., Witham, UK) and water ad libitum. Surgical procedures Rats were initially anaesthetized in an induction chamber with 4^5% iso£urane (vaporizer setting) delivered in oxygen (2 L minute1). Once unconscious (based on loss of righting re£exes and changes in respiratory rate), rats were transferred to a face mask system with coaxial scavenging and placed on a homeothermic blanket system to maintain body temperature (Harvard Apparatus, Edenbridge, UK). Anaesthesia was maintained with 1.5^2.5% iso£urane (vaporizer setting) delivered in oxygen. Ophthalmic ointment (Chloromycetin, Parke^Davis, Martindale Pharmaceuticals Ltd., Romford, UK) was applied to prevent corneal desiccation or abrasions. The pedal withdrawal response (PWR) was the primary method for assessing depth of anaesthesia. The strength of the re£ex was determined by extending the leg and pinching the interdigital web of the foot using the nails of the fore¢nger and thumb. The responses obtained were scored on a scale of 1^5, 1 indicating no response and 5 a marked response (Table 1). Respiratory rate, rhythm and pattern, eye blinking and whisker movements were also used in the overall evaluation of depth of anaesthesia. An attempt was made to minimize variation in the subjective assessment of anaesthetic depth using the aforementioned criteria, by ensuring that a single experimenter generated the stimuli and assessed all re£exes. Haemostats were not used to generate noxious stimulation because they damaged tissue when the ratchets engaged. Furthermore, it was felt that their use was equally subjective and o¡ered poor control of stimulus intensity. A 24 SWG intravenous cannula was inserted into the lateral tail vein and £ushed with heparinized saline. A midline subcutaneous injection of 0.2 mL of 1% lignocaine (Astra Pharmaceutical Ltd., Kings Langley, UK) was given in the scalp. Then, 23 minutes later, the skull was exposed by a 2-cm midline incision in the scalp and the right temporalis muscle was resected laterally. The exposed periosteum was cleared and the cranium cleaned with gauze swabs. Electrodes were silver/silver chloride balls (0.5 mm diameter) soldered to 0.7 mm diameter PVC insulated multicore leads. Two electrodes were chronically implanted through apertures bored in the cranium at the near vertex (Vx) and over the auditory cortex 197
Depth of anaesthesia monitoring with AEP LM Antunes et al.
Table 1 Clinical signs used to evaluate depth of anaesthesia
Depth
Symptoms
1 2
Lack of any response to PWR stimulus, slow and deep respirations Slight rise in muscle tone in response to the PWR stimulus considered to be a superficial reflex response Slight withdrawal response confined to the pinched limb, considered to be a stronger reflex response and related with ‘lighter’ anaesthesia Obvious response in tested leg. Occasional movement elsewhere in response to PWR and increased respiratory rate Onset of slight whisker movement or eye blinking after corneal stimulation, rapid withdrawal of leg and spontaneous movement of other limbs in response to PWR
3 4 5
Figure 1 Electrode localization on the rat skull. Over the auditory cortex (Acx), at the near vertex (Vx) and the reference (Ref).
probe was placed across the hind-foot to record arterial O2 saturation and heart rate. Electrode leads were soldered to a Jermyn 14-way DIL pin socket for connection to preampli¢ers (World Precision Instruments, Stevenage, UK) (gain 1000). Signals were band-pass 1^300 Hz ¢ltered (Neurolog NL125; Hertfordshire, UK), and sequenced to a CED Micro 1401 A/D converter (Cambridge Electronic Design Ltd, Cambridge, UK) for averaging with a microcomputer installed with CED (Signal) software. Stimuli were 100 clicks presented at 2 Hz for recording averaged AEP responses. Responses were sampled at 3000 Hz and recorded for 50 ms prior to each stimulus and 450 ms following presentation. Data collection and anaesthesia
(Acx) so that they made gentle contact with the cortical surface. (Fig. 1). The near vertex electrode was located 5^6 mm anterior to the interaural line and 1 mm lateral to the central suture. The auditory cortex electrode was 4^5 mm anterior to the interaural line, and 4^5 mm ventral to the dorsal skull on the right side (Simpson & Knight1993). A third electrode, the reference (Ref) was located over the frontal sinus. Electrodes were ¢xed with methyl methacrylate activated powder (Simplex Rapid, Associated Dental Products Ltd, Wiltshire, UK). The preparation was earthed with a length of ¢ne copper wire sutured to the skin overlying the nape of the neck.
Iso£urane administration was discontinued and the animals allowed to partly recover from anaesthesia (PWR score 4^5). An initial dose of fentanyl/£uanisone^midazolam (fentanyl 16 mg kg1/£uanisone 0.5 mg kg1^midazolam 0.25 mg kg1) was given intravenously (IV) (time ¼ 0). This was followed by further incremental injections 2, 4, 8 and 12 times this dose given at 30 minute intervals following initial injection. Auditory evoked potential data, arterial O2 saturation, respiratory rate and PWR were recorded at 5 minutes (deep anaesthesia) and 25 minutes (lighter anaesthesia) during each successive 30 minute period after drug injection. This is illustrated in Fig. 2.
Recording apparatus Animals were placed in a sound-attenuated box and held in sternal recumbency within a stereotaxic frame.Two hollow ear bars each ¢xed to a stereo loudspeaker (Sony, Japan) were used to deliver auditory (broad band click) stimuli. Clicks were determined to be 105 db and of 0.1 ms duration. A pulse-oximeter 198
Data analysis Analyses of the waveform components of the averaged AEP data obtained following each drug injection identi¢ed ¢ve characteristic peaks occurring at 11,14,23,42 and 68 milliseconds.These are illustrated in Fig. 3. The AEP waveform recorded at Acx was Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
Depth of anaesthesia monitoring with AEP LM Antunes et al.
Figure 2 Graphical representation of the experimental protocol. The ¢rst dose (fentanyl 16 mg kg1/£uanisone 0.5 mg kg1midazolam 0.25 mg kg1) administered at time ¼ 0 was followed by incremental doses 2, 4, 8 and 12 times dose 1 at 30-minute intervals. Data were collected 5 and 25 minutes after anaesthetic administration.
Figure 3 Graphical representation of the theoretical auditory evoked potential in the rat during hypnorm-midazolam anaesthesia with mean latency times for the observed peaks (P ¼ positive, N ¼ negative).
Figure 4 Rat 4 before (time 25 and 55 minutes) and after 0.5 mg kg1 midazolam with 32 mg kg1 fentanyl/1 mg kg1 £uanisone, and 1 mg kg1 midazolam with 64 mg kg1 fentanyl/2 mg kg1 £uanisone administrations (time 35 and 65 minutes). There is a delay in the peak times, the waveform moved to the right in the time scale after drug administration. These responses appeared with an increase in amplitude. Latency times from peaks P14, N23, P42 (ms) and amplitude N23^P42 (mV) are shown adjacent to the waveform. Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
199
Depth of anaesthesia monitoring with AEP LM Antunes et al.
essentially triphasic, consisting of one major positive peak (P42) and two smaller peaks (P14 and N23). Responses obtained at the Vx recording site were more inconsistent and less clearly de¢ned than those at Acx. Consequently, all analyses used data from Acx peaks P14, N23 and P42. The AEP data were measured according to peak amplitude and latency. Latency was the time of stimulus onset to the maximum amplitude of each main peak in the averaged response waveform. Amplitude was the di¡erence between the maximum negative de£ection (peak N23) and the maximum positive de£ection (peak P42; Figs. 3 and 4). The amplitude and latency of each AEP component and PWR were calculated during periods when anaesthesia was assumed to be ‘deep’ (5 minutes post injection) and ‘light’ (25 minutes post injection). All statistical analyses were performed using the software package SPSS 9.0 (SPSS, Chicago, IL, USA). Repeated measures ANOVA was used to determine the overall e¡ects of progressively increasing anaesthetic dose. For these analyses drug dose was used as the within subjects factor for comparing means comprising all data calculated for each type of assessment (PWR and individual peak latencies and amplitude) between successive sampling times. Individual paired t-tests were used to determine any change in depth of anaesthesia between recordings made 5 minutes following injection and those 20 minutes later. As the latter analyses involved ¢ve separate paired comparisons for withdrawal scores and each AEP component, a Bonferroni adjustment of probability level was used such that only values of p 0.01 were considered signi¢cant. To determine any relationship between changes in each AEP component and PWR, bivariate correlation analysis (Pearson coe⁄cient) was used to compare the overall mean change in latency of each AEP component and similarly calculated changes in PWR.
Averaged univariate results are quoted in the form; avg. F(df1, df2) ¼ x, p ¼ sig., where df1 and df2 are the degrees of freedom for the numerator and denominator, respectively. The results of paired t-tests are presented as; t(df) ¼ x, p ¼ sig. and Pearson correlations as r ¼ x, p ¼ sig.
Results The mean arterial oxygen saturation for all animals and for all time points ( SD) was 85 2.1%. A short period of apnoea was observed in each animal immediately following the ¢rst bolus of anaesthetic, but thereafter respiratory rate remained within a clinically acceptable range throughout anaesthesia (58 6 breaths minute1). Mean calculations ( SD) for all data collected following each successive dose of anaesthetic are shown in Table 2. Analysis of these data found a small reduction in PWR strength with successive and increasing anaesthetic dosage [avg. F(1,4) ¼ 3.07, p ¼ 0.047]. Pedal withdrawal response showed evidence of recovery during the 20 minutes following each injection, however, this was only signi¢cant for the 20 minutes following ¢rst administration of fentanyl/£uanisone^midazolam [t(3) ¼ 7.35, p ¼ 0.005]. Auditory evoked potential peaks P14, N23 and P42 all showed a signi¢cant increase in latency with successive injections of anaesthetic [avg. F(1,4) ¼ 12.53, p < 0.001; avg. F(1,4) ¼ 10.6, p < 0.001; avg. F(1,4) ¼ 3.9, p ¼ 0.02, respectively]. Peak N23 showed a signi¢cant reduction in latency during the 20 minute recovery following both the ¢rst and second bolus injection [t(3) ¼ 7.52, p ¼ 0.005; t(4) ¼ 5.17, p ¼ 0.007, respectively], but thereafter showed no signi¢cant change in peak latency. Peak P42 showed no signi¢cant change in latency during the 20 minutes following ¢rst drug administration (p ¼ 0.07), but occurred signi¢cantly earlier 20 minutes
Table 2 Latency times and pedal response (PWR) score in each interval of data collection following anaesthetic administration. Mean ( SD)
Time (minutes)
PWR
P14
N23
P42
N
0–30 30–60 60–90 90–120 120–150
2.2 0.76 1.9 0.9 1.1 0.22 1.4 0.9 10
15.5 0.7 15.4 2.3 17.1 2.8 19.1 2.2 21.8 3.0
26.3 3.4 24.9 3.0 28 2.3 29.9 1.6 34.7 4.8
45.6 4.8 44.4 3.4 47.1 2.3 48.8 3.9 51.6 2.3
5 5 5 5 5
200
Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
Depth of anaesthesia monitoring with AEP LM Antunes et al.
Figure 5 Changes in amplitudes before (time 25 and 55 minutes) and after (time 35 and 65 minutes) administration of 0.5 mg kg1 midazolam with 32 mg kg1 fentanyl/1 mg kg1 £uanisone, and 1 mg kg1 midazolam with 64 mg kg1 fentanyl/2 mg kg1 £uanisone), respectively.
following the second (2 initial dosage) anaesthetic injection [t(4) ¼ 4.75, p ¼ 0.009]. The mean overall depth of anaesthesia assessed using PWR scores was signi¢cantly correlated with the mean latency of peak N23, such that as the strength of PWR increased, N23 occurred signi¢cantly earlier (r ¼ 0.99, p ¼ 0.01). To determine the e¡ect of increasing the anaesthetic concentration, the AEP amplitude di¡erence (N23 to P42) was compared between data obtained at 25 minutes following each injection (light anaesthesia) and 5 minutes following the next bolus of anaesthetic (10 minute interval) (Figs 4 and 5).These showed increases in amplitude resulting from drug administration on both the second and third occasions [avg. F(1,4) ¼ 10.65, p ¼ 0.031; avg. F(1,4) ¼ 11.24, p ¼ 0.028, respectively].
Discussion Neuromuscular blocking agents suppress somatosensory responses and so the depth of anaesthesia must be assessed using variables such as heart rate and blood pressure when these drugs are used. However, these are of little value when fentanyl/£uanisone^midazolam is used in rats (Whelan 1996). Our study aimed to examine the potential of averaged mid-latency AEP responses as an anaesthetic ‘depth’ assessment technique. It was shown that progressive increases in anaesthetic dosage signi¢cantly reduced the overall strength of PWR and increased the latency of three peaks in averaged AEP responses (P14, N23 and P42). Moreover, the latency of two of these peaks, N23 and P42, indicated sensitivity to depth of anaesthesia during the 20 minutes prior to doubling the dose of anaesthetic given. The latency of peak N23 closely correlated with assessments of Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
PWR and so this component may be particularly useful for assessing the depth of anaesthesia produced by this, and possibly other, anaesthetics. Notable among our data was an increase in the AEP amplitudes which followed incremental dosing, because most anaesthetics produce the opposite e¡ect in humans (Thornton et al. 1989; Schwender et al. 1995, 1997) and rats (Jensen et al. 1998). These latter authors reported that calculation of an index comprising all characteristics of the AEP response correlated well with loss of somatosensory responses. Comparisons between our study and the ¢ndings of Jensen et al. (1998) cannot be made because the method of index calculation precludes identifying the contribution of di¡erent AEP components. In addition, Jensen et al. (1998) gave incremental doses at 5 minute intervals; which may have meant that there was inadequate time to assess transitions to lighter anaesthetic levels. A recent investigation that studied induction and recovery from anaesthesia induced with the same agents (given by intramuscular and intraperitoneal injection) con¢rmed the usefulness of the mid latency AEP record and Vertex location as an indicator of hypnosis, although unexpectedly, an increase in the N22^P34 interpeak was observed immediately after drug administration (Haberham et al. 2000). The e¡ect of this anaesthetic combination on the AEP in humans has not been studied, although midazolam and fentanyl have been assessed independently, with con£icting results. According to Brunner et al. (1999) loss of consciousness following midazolam injection in man is associated with increased AEP latency, although Schwender et al. (1993a) found no such changes. However, endotracheal intubation was performed in the latter, but not in the former study once consciousness was lost. It is 201
Depth of anaesthesia monitoring with AEP LM Antunes et al.
possible that somatosensory stimulation generated by this act reversed the normally depressant e¡ect, i.e. reduced amplitude and, or increased latency (Richmond et al. 1996) of midazolam on AEPs. Schwender et al. (1993b) reported no AEP alterations in response to incremental doses of alfentanil, fentanyl and morphine but this is similarly at odds with more recent work (Crabb et al. 1996) in which remifentanil caused amplitude suppression in a doserelated manner. Reports that diazepam or fentanyl have no e¡ect upon brainstem, visual or somatosensory evoked responses in man (Loughnan et al. 1987), con¢rm the variable e¡ects of opioids and benzodiazepines upon AEP characteristics. This is in contrast with the uniformly depressant e¡ects of inhalational anaesthetics, and makes it di⁄cult to formulate a convincing hypothesis for the encountered increases in amplitude following drug treatment. Nevertheless, such a hypothesis might involve anaesthetic-induced inhibition of some of the neuronal pathways normally involved in sensory depression, and further studies are underway to determine which components of this anaesthetic combination are principally involved. Although the latency changes observed in this study, which we attribute to drug e¡ects, are consistent with previous observations, it is possible that the periods of apnoea following drug administration may have been in£uential, as hypercapnia and hypoxaemia are known to a¡ect AEP latency. However, we consider this unlikely as the apnoeic periods were very brief and occurred only after the initial injection. Furthermore, work in cats indicates that hypercapnia does not signi¢cantly a¡ect auditory brainstem recordings (ABR) unless severe (Freeman et al.1991). We conclude that the latency time of some peaks of the AEP may be a useful tool for evaluating depth of anaesthesia in animals, in particularly peaks N (23) and P (42). Further studies are needed: (i) to determine which component of the anaesthetic used in this study exerted the greatest in£uence on AEP latency and amplitude changes; and (ii) to evaluate the e¡ect of di¡erent anaesthetics on the consistency of these changes. It should be noted that dural electrodes were used in this study in order to maximize signal strength, minimize noise and allow the relatively rapid determination of AEP characteristics. This technique is only appropriate for nonsurvival experiments. In other circumstances, consideration for animal comfort would warrant the use of subdermal electrodes, 202
in which case a greater number of clicks would need to be averaged to maintain signal quality.
Acknowledgements Luis Antunes is sponsored by a PRAXIS XXI Grant, Science and Technology Foundation, Lisbon.
References Brunner MD, Umo-Etuk J, Sharpe RM et al. (1999) E¡ect of a bolus dose of midazolam on the auditory evoked response in humans. Br J Anaesth 82,633^634. Bullock TH (1981) Neuroethology deserves more study of evoked responses. Neuroscience 6,1203^1215. Crabb I,Thornton C, Konieczko KM et al. (1996) Remifentanil reduces auditory and somatosensory evoked responses during iso£urane anaesthesia in a dose-dependent manner. Br J Anaesth 76,795^801. Flecknell PA, Mitchell M (1984) Midazolam and fentanyl^ £uanisone: assessment of anaesthetic e¡ects in laboratory rodents and rabbits. Lab Anim18,143^146. Fox SS, O’Brian JH (1965) Duplication of evoked potential waveform by curve of probability of ¢ring a single cell. Science147,888^890. Freeman S, Sohmer H, Silver S (1991) The e¡ect of stimulus repetition rate on the diagnostic e⁄cacy of the auditory nervebrain-stem evoked response. Electroencephalogr Clin Neurophysiol 78, 284^290. Haberham ZL, van den Brom WE, Venker-van Haagen AJ et al. (1999) EEG evaluation of re£ex testing as assessment of depth of pentobarbital anaesthesia in the rat. LabAnim 33, 47^57. Haberham ZL, Alfarez DN, van de Brom WE et al. (2000) Di¡erential monitoring of hypnosis and anti-nociception during fentanyl/£uanisone/midazolam anaesthesia using spontaneous and evoked electroencephalography in the rat. In: Haberham ZL (eds). Development and Evaluation of Methods for Assessment of Quality of Anaesthesia in the Rat. PhD Thesis. Universiteit Utrecht, Utrecht, pp.125^146. Jensen EW, Nygaard M, Henneberg SW (1998) On-line analysis of middle latency auditory evoked potentials (MLAEP) for monitoring depth of anaesthesia in laboratory rats. Med Eng Phys 20,722^728. Loughnan BL, Sebel PS,Thomas D et al. (1987) Evoked potentials following diazepam or fentanyl. Anaesthesia 42, 195^198. Rampil IJ, Laster MJ (1992) No correlation between quantitative electroencephalographic measurements and movement response to noxious stimuli during iso£urane anesthesia in rats. Anesthesiology 77,920^925. Richmond CE, Matson A, Thornton C et al. (1996) E¡ect of neuromuscular block on depth of anaesthesia as measured by the auditory evoked response. Br J Anaesth 76, 446^448. Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
Depth of anaesthesia monitoring with AEP LM Antunes et al.
Schwender D, Klasing S, Madler C et al. (1993a) E¡ects of benzodiazepines on mid-latency auditory evoked potentials. Can J Anaesth 40,1148^1154. Schwender D, Rimkus T, Haessler R et al. (1993b) E¡ects of increasing doses of alfentanil, fentanyl and morphine on mid-latency auditory evoked potentials. Br J Anaesth 71, 622^628. Schwender D, Madler C, Klasing S et al. (1995) Mid-latency auditory evoked potentials and wakefulness during Caesarean section. Eur J Anaesthesiol12,171^179. Schwender D, Daunderer M, Mulzer S et al. (1997) Midlatency auditory evoked potentials predict movements during anesthesia with iso£urane or propofol. Anesth Analg 85,164^173. Simpson GV, Knight RT (1993) Multiple brain systems generating the rat auditory evoked potential I. Character-
Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
ization of the auditory cortex response. Brain Res 602, 240^250. Thornton C (1991) Evoked potentials in anaesthesia. Eur J Anaesthesiol 8,89^107. Thornton C, Konieczko KM, Knight AB et al. (1989) E¡ect of propofol on the auditory evoked response and oesophageal contractility. Br J Anaesth 63, 411^417. Whelan G (1996) The assessment of depth of anaesthesia and the e¡ects of anaesthesia in the rat (Rattus norvegicus). PhD Thesis, Newcastle upon Tyne. Whelan G, Flecknell PA (1992) The assessment of depth of anaesthesia in animals and man. Lab Anim 26, 153^162.
Received 7 November 2000; accepted 3 February 2001.
203
R E S E A R C H PA P E R
Evaluation of auditory evoked potentials to predict depth of anaesthesia during fentanyl/fluanisonemidazolam anaesthesia in rats LM Antunes
MSc, DVM,
JV Roughan
BSc, PhD
& PA Flecknell
MA, Vet MB, PhD, DLAS, Dip ECVA
Comparative Biology Centre, Medical School, University of Newcastle upon Tyne, UK
Correspondence: Luis Antunes, Comparative Biology Centre, Medical School, University of Newcastle upon Tyne NE2 4HH, UK.
Abstract Objectives To assess a method for monitoring depth of anaesthesia using components of middle latency auditory evoked potential (AEP) waveforms during anaesthesiawith fentanyl/£uanisoneand midazolam. Study design Prospective observational study. Animals Five female Wistar rats weighing between 210 and 250 g. Methods Implanted electrodes were used to record AEPs in animals receiving ¢ve doses of anaesthetic. Recordings were made at 5 minutes post-injection (deep anaesthesia; no pedal withdrawal response, PWR) and then at 25 minutes (light anaesthesia; strong PWR). Responses showed ¢ve characteristic peaks occurring at 11, 14, 23, 42 and 68 ms that were measured for latency of occurrence and peak amplitude. Results Auditory evoked potential peaks P14, N23 and P42 were increased signi¢cantly in latency with successive anaesthetic injections [avg. F(1,4) ¼ 12.53, p < 0.001; avg. F(1,4) ¼ 10.6, p < 0.001; avg. F(1,4) ¼ 3.9, p ¼ 0.02, respectively]. Peak N23 showed a signi¢cant reduction in latency during the 20 minute recovery period following both the ¢rst and second anaesthetic injections (t(3) ¼ 7.52, p ¼ 0.005; t(4) ¼ 5.17, p ¼ 0.007, respectively). Peak P42 occurred signi¢cantly earlier 20 minutes following the second anaesthetic injection (t(4) ¼ 4.75, p ¼ 0.009). The mean overall depth of anaesthesia assessed using PWR scores was signi¢cantly correlated with the mean latency of peak N23, such that as the strength of PWR increased, N23 occurred 196
signi¢cantly earlier (r ¼ 0.99, p ¼ 0.01). The amplitude di¡erence between peaks N23 and P42 increased after the second and third drug administrations [avg. F(1,4) ¼ 10.65, p ¼ 0.031 and avg. F(1,4) ¼ 11.24, p ¼ 0.028, respectively]. Conclusion The characteristics of these peaks, and in particular latency of peak N23, may provide a useful tool for assessing depth of anaesthesia produced by this, and possibly other anaesthetic agents. Keywords anaesthesia, auditory evoked potential, fentanyl/ £uanisone/ midazolam, rat.
Introduction The Animals (Scienti¢c Procedures) Act (1986) and European Directive (86/609/EEC) de¢ne the care requirements of experimental animals undergoing surgical procedures and impose a greater obligation on researchers to ensure the provision of adequate anaesthesia. Several recent studies in humans have considered changes in auditory evoked potential (AEP) activity as a potentially useful tool for evaluating anaesthetic depth (Thornton 1991). Auditory evoked responses are of short (approximately <100 millisecond) duration, and are generated as an average of EEG activity following repetitive stimulus presentation. Their amplitude and latency are thought to re£ect the activation state of neuronal populations (Fox & O’Brian 1965). Evoked responses are considered powerful diagnostic tools as they are frequently closely correlated with behavioural, cognitive and clinical states (Bullock 1981). Auditory evoked responses are enhanced in amplitude and
Depth of anaesthesia monitoring with AEP LM Antunes et al.
reduced in latency when general arousal levels are high, contrary to the circumstances encountered during anaesthesia. Accordingly, dose-dependent amplitude depression and increased latency of AEP components during anaesthesia are now generally accepted indices of depth for some anaesthetics (e.g. volatile anaesthetics and propofol) (Thornton et al. 1989; Schwender et al. 1995, 1997). Relatively few animal studies have been conducted, and of these, most have assessed techniques such as respiratory rate, loss of the righting re£ex and strength of pedal withdrawal response (PWR) (Whelan & Flecknell 1992). This may be due, in part, to reports of poor correlation between EEG parameters and the strength of these somatic responses, for example during pentobarbital anesthesia (Haberham et al.1999) and iso£urane anaesthesia in rats (Rampil & Laster1992). Because of the increasing use of anaesthetic techniques employing neuromuscular blocking agents in which these types of assessment are unavailable, there is now a requirement for a more reliable and objective assessment technique. Previous studies used a broad spectrum of anaesthetic agents across a range of species. The con£icting data produced may have resulted from the lack of a su⁄ciently systematic approach to anaesthesia depth assessment; thus, AEP measurements may represent a useful alternative. Here we describe the results of a pilot study to characterize the average EEG response of rats to a series of click stimuli, delivered during anaesthesia induced with iso£urane and subsequently maintained with increasing doses of fentanyl/£uanisone^midazolam, a neuroleptanalgesic combination commonly used to anaesthetize laboratory rats (Flecknell & Mitchell1984).
Materials and methods All procedures were authorized under the Animals (Scienti¢c Procedures) Act (1986) and were approved by the institutional ethics committee. Five female outbred Wistar rats (Charles River, UK) weighing between 210 and 250 g and free from recognized respiratory pathogens were used. Animals were grouphoused in solid-£oored caging (RC1, North Kent Plastics Ltd., Erith, UK;56 cm 38 cm, height18 cm). Sawdust (‘Gold Chip’, BS and S Ltd., Edinburgh, UK) was used as bedding. Stocking densities were within UK-speci¢ed guidelines [Animals (Scienti¢c Procedures) Act 1986; 1989]. Room temperature was controlled at 20 2 8C. A 9:15 hours light/dark cycle was maintained (lights on 08:00 hours o¡ 17 : 00 hours). Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
Animals received a commercial pellet diet (R and M no. 1, Special Diets Services Ltd., Witham, UK) and water ad libitum. Surgical procedures Rats were initially anaesthetized in an induction chamber with 4^5% iso£urane (vaporizer setting) delivered in oxygen (2 L minute1). Once unconscious (based on loss of righting re£exes and changes in respiratory rate), rats were transferred to a face mask system with coaxial scavenging and placed on a homeothermic blanket system to maintain body temperature (Harvard Apparatus, Edenbridge, UK). Anaesthesia was maintained with 1.5^2.5% iso£urane (vaporizer setting) delivered in oxygen. Ophthalmic ointment (Chloromycetin, Parke^Davis, Martindale Pharmaceuticals Ltd., Romford, UK) was applied to prevent corneal desiccation or abrasions. The pedal withdrawal response (PWR) was the primary method for assessing depth of anaesthesia. The strength of the re£ex was determined by extending the leg and pinching the interdigital web of the foot using the nails of the fore¢nger and thumb. The responses obtained were scored on a scale of 1^5, 1 indicating no response and 5 a marked response (Table 1). Respiratory rate, rhythm and pattern, eye blinking and whisker movements were also used in the overall evaluation of depth of anaesthesia. An attempt was made to minimize variation in the subjective assessment of anaesthetic depth using the aforementioned criteria, by ensuring that a single experimenter generated the stimuli and assessed all re£exes. Haemostats were not used to generate noxious stimulation because they damaged tissue when the ratchets engaged. Furthermore, it was felt that their use was equally subjective and o¡ered poor control of stimulus intensity. A 24 SWG intravenous cannula was inserted into the lateral tail vein and £ushed with heparinized saline. A midline subcutaneous injection of 0.2 mL of 1% lignocaine (Astra Pharmaceutical Ltd., Kings Langley, UK) was given in the scalp. Then, 23 minutes later, the skull was exposed by a 2-cm midline incision in the scalp and the right temporalis muscle was resected laterally. The exposed periosteum was cleared and the cranium cleaned with gauze swabs. Electrodes were silver/silver chloride balls (0.5 mm diameter) soldered to 0.7 mm diameter PVC insulated multicore leads. Two electrodes were chronically implanted through apertures bored in the cranium at the near vertex (Vx) and over the auditory cortex 197
Depth of anaesthesia monitoring with AEP LM Antunes et al.
Table 1 Clinical signs used to evaluate depth of anaesthesia
Depth
Symptoms
1 2
Lack of any response to PWR stimulus, slow and deep respirations Slight rise in muscle tone in response to the PWR stimulus considered to be a superficial reflex response Slight withdrawal response confined to the pinched limb, considered to be a stronger reflex response and related with ‘lighter’ anaesthesia Obvious response in tested leg. Occasional movement elsewhere in response to PWR and increased respiratory rate Onset of slight whisker movement or eye blinking after corneal stimulation, rapid withdrawal of leg and spontaneous movement of other limbs in response to PWR
3 4 5
Figure 1 Electrode localization on the rat skull. Over the auditory cortex (Acx), at the near vertex (Vx) and the reference (Ref).
probe was placed across the hind-foot to record arterial O2 saturation and heart rate. Electrode leads were soldered to a Jermyn 14-way DIL pin socket for connection to preampli¢ers (World Precision Instruments, Stevenage, UK) (gain 1000). Signals were band-pass 1^300 Hz ¢ltered (Neurolog NL125; Hertfordshire, UK), and sequenced to a CED Micro 1401 A/D converter (Cambridge Electronic Design Ltd, Cambridge, UK) for averaging with a microcomputer installed with CED (Signal) software. Stimuli were 100 clicks presented at 2 Hz for recording averaged AEP responses. Responses were sampled at 3000 Hz and recorded for 50 ms prior to each stimulus and 450 ms following presentation. Data collection and anaesthesia
(Acx) so that they made gentle contact with the cortical surface. (Fig. 1). The near vertex electrode was located 5^6 mm anterior to the interaural line and 1 mm lateral to the central suture. The auditory cortex electrode was 4^5 mm anterior to the interaural line, and 4^5 mm ventral to the dorsal skull on the right side (Simpson & Knight1993). A third electrode, the reference (Ref) was located over the frontal sinus. Electrodes were ¢xed with methyl methacrylate activated powder (Simplex Rapid, Associated Dental Products Ltd, Wiltshire, UK). The preparation was earthed with a length of ¢ne copper wire sutured to the skin overlying the nape of the neck.
Iso£urane administration was discontinued and the animals allowed to partly recover from anaesthesia (PWR score 4^5). An initial dose of fentanyl/£uanisone^midazolam (fentanyl 16 mg kg1/£uanisone 0.5 mg kg1^midazolam 0.25 mg kg1) was given intravenously (IV) (time ¼ 0). This was followed by further incremental injections 2, 4, 8 and 12 times this dose given at 30 minute intervals following initial injection. Auditory evoked potential data, arterial O2 saturation, respiratory rate and PWR were recorded at 5 minutes (deep anaesthesia) and 25 minutes (lighter anaesthesia) during each successive 30 minute period after drug injection. This is illustrated in Fig. 2.
Recording apparatus Animals were placed in a sound-attenuated box and held in sternal recumbency within a stereotaxic frame.Two hollow ear bars each ¢xed to a stereo loudspeaker (Sony, Japan) were used to deliver auditory (broad band click) stimuli. Clicks were determined to be 105 db and of 0.1 ms duration. A pulse-oximeter 198
Data analysis Analyses of the waveform components of the averaged AEP data obtained following each drug injection identi¢ed ¢ve characteristic peaks occurring at 11,14,23,42 and 68 milliseconds.These are illustrated in Fig. 3. The AEP waveform recorded at Acx was Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
Depth of anaesthesia monitoring with AEP LM Antunes et al.
Figure 2 Graphical representation of the experimental protocol. The ¢rst dose (fentanyl 16 mg kg1/£uanisone 0.5 mg kg1midazolam 0.25 mg kg1) administered at time ¼ 0 was followed by incremental doses 2, 4, 8 and 12 times dose 1 at 30-minute intervals. Data were collected 5 and 25 minutes after anaesthetic administration.
Figure 3 Graphical representation of the theoretical auditory evoked potential in the rat during hypnorm-midazolam anaesthesia with mean latency times for the observed peaks (P ¼ positive, N ¼ negative).
Figure 4 Rat 4 before (time 25 and 55 minutes) and after 0.5 mg kg1 midazolam with 32 mg kg1 fentanyl/1 mg kg1 £uanisone, and 1 mg kg1 midazolam with 64 mg kg1 fentanyl/2 mg kg1 £uanisone administrations (time 35 and 65 minutes). There is a delay in the peak times, the waveform moved to the right in the time scale after drug administration. These responses appeared with an increase in amplitude. Latency times from peaks P14, N23, P42 (ms) and amplitude N23^P42 (mV) are shown adjacent to the waveform. Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
199
Depth of anaesthesia monitoring with AEP LM Antunes et al.
essentially triphasic, consisting of one major positive peak (P42) and two smaller peaks (P14 and N23). Responses obtained at the Vx recording site were more inconsistent and less clearly de¢ned than those at Acx. Consequently, all analyses used data from Acx peaks P14, N23 and P42. The AEP data were measured according to peak amplitude and latency. Latency was the time of stimulus onset to the maximum amplitude of each main peak in the averaged response waveform. Amplitude was the di¡erence between the maximum negative de£ection (peak N23) and the maximum positive de£ection (peak P42; Figs. 3 and 4). The amplitude and latency of each AEP component and PWR were calculated during periods when anaesthesia was assumed to be ‘deep’ (5 minutes post injection) and ‘light’ (25 minutes post injection). All statistical analyses were performed using the software package SPSS 9.0 (SPSS, Chicago, IL, USA). Repeated measures ANOVA was used to determine the overall e¡ects of progressively increasing anaesthetic dose. For these analyses drug dose was used as the within subjects factor for comparing means comprising all data calculated for each type of assessment (PWR and individual peak latencies and amplitude) between successive sampling times. Individual paired t-tests were used to determine any change in depth of anaesthesia between recordings made 5 minutes following injection and those 20 minutes later. As the latter analyses involved ¢ve separate paired comparisons for withdrawal scores and each AEP component, a Bonferroni adjustment of probability level was used such that only values of p 0.01 were considered signi¢cant. To determine any relationship between changes in each AEP component and PWR, bivariate correlation analysis (Pearson coe⁄cient) was used to compare the overall mean change in latency of each AEP component and similarly calculated changes in PWR.
Averaged univariate results are quoted in the form; avg. F(df1, df2) ¼ x, p ¼ sig., where df1 and df2 are the degrees of freedom for the numerator and denominator, respectively. The results of paired t-tests are presented as; t(df) ¼ x, p ¼ sig. and Pearson correlations as r ¼ x, p ¼ sig.
Results The mean arterial oxygen saturation for all animals and for all time points ( SD) was 85 2.1%. A short period of apnoea was observed in each animal immediately following the ¢rst bolus of anaesthetic, but thereafter respiratory rate remained within a clinically acceptable range throughout anaesthesia (58 6 breaths minute1). Mean calculations ( SD) for all data collected following each successive dose of anaesthetic are shown in Table 2. Analysis of these data found a small reduction in PWR strength with successive and increasing anaesthetic dosage [avg. F(1,4) ¼ 3.07, p ¼ 0.047]. Pedal withdrawal response showed evidence of recovery during the 20 minutes following each injection, however, this was only signi¢cant for the 20 minutes following ¢rst administration of fentanyl/£uanisone^midazolam [t(3) ¼ 7.35, p ¼ 0.005]. Auditory evoked potential peaks P14, N23 and P42 all showed a signi¢cant increase in latency with successive injections of anaesthetic [avg. F(1,4) ¼ 12.53, p < 0.001; avg. F(1,4) ¼ 10.6, p < 0.001; avg. F(1,4) ¼ 3.9, p ¼ 0.02, respectively]. Peak N23 showed a signi¢cant reduction in latency during the 20 minute recovery following both the ¢rst and second bolus injection [t(3) ¼ 7.52, p ¼ 0.005; t(4) ¼ 5.17, p ¼ 0.007, respectively], but thereafter showed no signi¢cant change in peak latency. Peak P42 showed no signi¢cant change in latency during the 20 minutes following ¢rst drug administration (p ¼ 0.07), but occurred signi¢cantly earlier 20 minutes
Table 2 Latency times and pedal response (PWR) score in each interval of data collection following anaesthetic administration. Mean ( SD)
Time (minutes)
PWR
P14
N23
P42
N
0–30 30–60 60–90 90–120 120–150
2.2 0.76 1.9 0.9 1.1 0.22 1.4 0.9 10
15.5 0.7 15.4 2.3 17.1 2.8 19.1 2.2 21.8 3.0
26.3 3.4 24.9 3.0 28 2.3 29.9 1.6 34.7 4.8
45.6 4.8 44.4 3.4 47.1 2.3 48.8 3.9 51.6 2.3
5 5 5 5 5
200
Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
Depth of anaesthesia monitoring with AEP LM Antunes et al.
Figure 5 Changes in amplitudes before (time 25 and 55 minutes) and after (time 35 and 65 minutes) administration of 0.5 mg kg1 midazolam with 32 mg kg1 fentanyl/1 mg kg1 £uanisone, and 1 mg kg1 midazolam with 64 mg kg1 fentanyl/2 mg kg1 £uanisone), respectively.
following the second (2 initial dosage) anaesthetic injection [t(4) ¼ 4.75, p ¼ 0.009]. The mean overall depth of anaesthesia assessed using PWR scores was signi¢cantly correlated with the mean latency of peak N23, such that as the strength of PWR increased, N23 occurred signi¢cantly earlier (r ¼ 0.99, p ¼ 0.01). To determine the e¡ect of increasing the anaesthetic concentration, the AEP amplitude di¡erence (N23 to P42) was compared between data obtained at 25 minutes following each injection (light anaesthesia) and 5 minutes following the next bolus of anaesthetic (10 minute interval) (Figs 4 and 5).These showed increases in amplitude resulting from drug administration on both the second and third occasions [avg. F(1,4) ¼ 10.65, p ¼ 0.031; avg. F(1,4) ¼ 11.24, p ¼ 0.028, respectively].
Discussion Neuromuscular blocking agents suppress somatosensory responses and so the depth of anaesthesia must be assessed using variables such as heart rate and blood pressure when these drugs are used. However, these are of little value when fentanyl/£uanisone^midazolam is used in rats (Whelan 1996). Our study aimed to examine the potential of averaged mid-latency AEP responses as an anaesthetic ‘depth’ assessment technique. It was shown that progressive increases in anaesthetic dosage signi¢cantly reduced the overall strength of PWR and increased the latency of three peaks in averaged AEP responses (P14, N23 and P42). Moreover, the latency of two of these peaks, N23 and P42, indicated sensitivity to depth of anaesthesia during the 20 minutes prior to doubling the dose of anaesthetic given. The latency of peak N23 closely correlated with assessments of Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
PWR and so this component may be particularly useful for assessing the depth of anaesthesia produced by this, and possibly other, anaesthetics. Notable among our data was an increase in the AEP amplitudes which followed incremental dosing, because most anaesthetics produce the opposite e¡ect in humans (Thornton et al. 1989; Schwender et al. 1995, 1997) and rats (Jensen et al. 1998). These latter authors reported that calculation of an index comprising all characteristics of the AEP response correlated well with loss of somatosensory responses. Comparisons between our study and the ¢ndings of Jensen et al. (1998) cannot be made because the method of index calculation precludes identifying the contribution of di¡erent AEP components. In addition, Jensen et al. (1998) gave incremental doses at 5 minute intervals; which may have meant that there was inadequate time to assess transitions to lighter anaesthetic levels. A recent investigation that studied induction and recovery from anaesthesia induced with the same agents (given by intramuscular and intraperitoneal injection) con¢rmed the usefulness of the mid latency AEP record and Vertex location as an indicator of hypnosis, although unexpectedly, an increase in the N22^P34 interpeak was observed immediately after drug administration (Haberham et al. 2000). The e¡ect of this anaesthetic combination on the AEP in humans has not been studied, although midazolam and fentanyl have been assessed independently, with con£icting results. According to Brunner et al. (1999) loss of consciousness following midazolam injection in man is associated with increased AEP latency, although Schwender et al. (1993a) found no such changes. However, endotracheal intubation was performed in the latter, but not in the former study once consciousness was lost. It is 201
Depth of anaesthesia monitoring with AEP LM Antunes et al.
possible that somatosensory stimulation generated by this act reversed the normally depressant e¡ect, i.e. reduced amplitude and, or increased latency (Richmond et al. 1996) of midazolam on AEPs. Schwender et al. (1993b) reported no AEP alterations in response to incremental doses of alfentanil, fentanyl and morphine but this is similarly at odds with more recent work (Crabb et al. 1996) in which remifentanil caused amplitude suppression in a doserelated manner. Reports that diazepam or fentanyl have no e¡ect upon brainstem, visual or somatosensory evoked responses in man (Loughnan et al. 1987), con¢rm the variable e¡ects of opioids and benzodiazepines upon AEP characteristics. This is in contrast with the uniformly depressant e¡ects of inhalational anaesthetics, and makes it di⁄cult to formulate a convincing hypothesis for the encountered increases in amplitude following drug treatment. Nevertheless, such a hypothesis might involve anaesthetic-induced inhibition of some of the neuronal pathways normally involved in sensory depression, and further studies are underway to determine which components of this anaesthetic combination are principally involved. Although the latency changes observed in this study, which we attribute to drug e¡ects, are consistent with previous observations, it is possible that the periods of apnoea following drug administration may have been in£uential, as hypercapnia and hypoxaemia are known to a¡ect AEP latency. However, we consider this unlikely as the apnoeic periods were very brief and occurred only after the initial injection. Furthermore, work in cats indicates that hypercapnia does not signi¢cantly a¡ect auditory brainstem recordings (ABR) unless severe (Freeman et al.1991). We conclude that the latency time of some peaks of the AEP may be a useful tool for evaluating depth of anaesthesia in animals, in particularly peaks N (23) and P (42). Further studies are needed: (i) to determine which component of the anaesthetic used in this study exerted the greatest in£uence on AEP latency and amplitude changes; and (ii) to evaluate the e¡ect of di¡erent anaesthetics on the consistency of these changes. It should be noted that dural electrodes were used in this study in order to maximize signal strength, minimize noise and allow the relatively rapid determination of AEP characteristics. This technique is only appropriate for nonsurvival experiments. In other circumstances, consideration for animal comfort would warrant the use of subdermal electrodes, 202
in which case a greater number of clicks would need to be averaged to maintain signal quality.
Acknowledgements Luis Antunes is sponsored by a PRAXIS XXI Grant, Science and Technology Foundation, Lisbon.
References Brunner MD, Umo-Etuk J, Sharpe RM et al. (1999) E¡ect of a bolus dose of midazolam on the auditory evoked response in humans. Br J Anaesth 82,633^634. Bullock TH (1981) Neuroethology deserves more study of evoked responses. Neuroscience 6,1203^1215. Crabb I,Thornton C, Konieczko KM et al. (1996) Remifentanil reduces auditory and somatosensory evoked responses during iso£urane anaesthesia in a dose-dependent manner. Br J Anaesth 76,795^801. Flecknell PA, Mitchell M (1984) Midazolam and fentanyl^ £uanisone: assessment of anaesthetic e¡ects in laboratory rodents and rabbits. Lab Anim18,143^146. Fox SS, O’Brian JH (1965) Duplication of evoked potential waveform by curve of probability of ¢ring a single cell. Science147,888^890. Freeman S, Sohmer H, Silver S (1991) The e¡ect of stimulus repetition rate on the diagnostic e⁄cacy of the auditory nervebrain-stem evoked response. Electroencephalogr Clin Neurophysiol 78, 284^290. Haberham ZL, van den Brom WE, Venker-van Haagen AJ et al. (1999) EEG evaluation of re£ex testing as assessment of depth of pentobarbital anaesthesia in the rat. LabAnim 33, 47^57. Haberham ZL, Alfarez DN, van de Brom WE et al. (2000) Di¡erential monitoring of hypnosis and anti-nociception during fentanyl/£uanisone/midazolam anaesthesia using spontaneous and evoked electroencephalography in the rat. In: Haberham ZL (eds). Development and Evaluation of Methods for Assessment of Quality of Anaesthesia in the Rat. PhD Thesis. Universiteit Utrecht, Utrecht, pp.125^146. Jensen EW, Nygaard M, Henneberg SW (1998) On-line analysis of middle latency auditory evoked potentials (MLAEP) for monitoring depth of anaesthesia in laboratory rats. Med Eng Phys 20,722^728. Loughnan BL, Sebel PS,Thomas D et al. (1987) Evoked potentials following diazepam or fentanyl. Anaesthesia 42, 195^198. Rampil IJ, Laster MJ (1992) No correlation between quantitative electroencephalographic measurements and movement response to noxious stimuli during iso£urane anesthesia in rats. Anesthesiology 77,920^925. Richmond CE, Matson A, Thornton C et al. (1996) E¡ect of neuromuscular block on depth of anaesthesia as measured by the auditory evoked response. Br J Anaesth 76, 446^448. Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
Depth of anaesthesia monitoring with AEP LM Antunes et al.
Schwender D, Klasing S, Madler C et al. (1993a) E¡ects of benzodiazepines on mid-latency auditory evoked potentials. Can J Anaesth 40,1148^1154. Schwender D, Rimkus T, Haessler R et al. (1993b) E¡ects of increasing doses of alfentanil, fentanyl and morphine on mid-latency auditory evoked potentials. Br J Anaesth 71, 622^628. Schwender D, Madler C, Klasing S et al. (1995) Mid-latency auditory evoked potentials and wakefulness during Caesarean section. Eur J Anaesthesiol12,171^179. Schwender D, Daunderer M, Mulzer S et al. (1997) Midlatency auditory evoked potentials predict movements during anesthesia with iso£urane or propofol. Anesth Analg 85,164^173. Simpson GV, Knight RT (1993) Multiple brain systems generating the rat auditory evoked potential I. Character-
Veterinary Anaesthesia and Analgesia, 2001, 28, 196^203
ization of the auditory cortex response. Brain Res 602, 240^250. Thornton C (1991) Evoked potentials in anaesthesia. Eur J Anaesthesiol 8,89^107. Thornton C, Konieczko KM, Knight AB et al. (1989) E¡ect of propofol on the auditory evoked response and oesophageal contractility. Br J Anaesth 63, 411^417. Whelan G (1996) The assessment of depth of anaesthesia and the e¡ects of anaesthesia in the rat (Rattus norvegicus). PhD Thesis, Newcastle upon Tyne. Whelan G, Flecknell PA (1992) The assessment of depth of anaesthesia in animals and man. Lab Anim 26, 153^162.
Received 7 November 2000; accepted 3 February 2001.
203