SOMATOSENSORY AND AUDITORY EVOKED RESPONSES RECORDED SIMULTANEOUSLY: DIFFERENTIAL EFFECTS OF NITROUS OXIDE AND TSOFLURANE

British Journal of Anaesthesia 1992; 68: 508-514

SOMATOSENSORY AND AUDITORY EVOKED RESPONSES RECORDED SIMULTANEOUSLY: DIFFERENTIAL EFFECTS OF NITROUS OXIDE AND TSOFLURANE C. THORNTON, P. CREAGH-BARRY, C. JORDAN, N. P. LUFF, C. J. DORE, M. HENLEY AND D. E. F. NEWTON

SUMMARY

KEY WORDS Anaesthetics, gases: nitrous oxide. Anaesthetics, volatile: isoflurane. Brain: auditory evoked potentials, somatosensory evoked potentials.

The measurement of depth of anaesthesia is a challenge to anaesthetists. The anaesthetic state encompasses drug-induced loss of consciousness or hypnosis, analgesia, amnesia, lack of movement in response to surgical incision and prevention of changes in autonomic signs. This series of functional deficits is arranged conventionally in a descending hierarchy to correspond with a concept of deepening anaesthesia. The problem is that the components of the hierarchy are not necessarily interdependent, and they are not affected to the same degree at equal MAC values [1] by various anaesthetic techniques. If

PATIENTS AND METHODS

We studied eight patients (five females) aged 26-47 yr and ASA I or II, who were undergoing major elective C. THORNTON, PH.D., P. CRBAGH-BARRY*, M.B., B.S., F.C.ANAES., C. JORDAN, M.SC., N. P. LUFF, B.SC., D. E. F. NEWTON, M.B., B.S.,

F.C.ANAES. (Division of Anaesthesia); C. J. DORE, B.SC., M. HENLEY, B.SC. (Section of Medical Statistics); Clinical Research Centre, Watford Road, Harrow, Middlesex HA1 3UJ. Accepted for Publication: November 28, 1991. •Present address: Department of Anaesthesia, Royal Free Hospital, Pond Street, London NW3.

Downloaded from http://bja.oxfordjournals.org/ at University of Ulster on June 25, 2015

Auditory (AER) and somatosensory evoked responses (SSER) were recorded simultaneously in eight patients under anaesthesia before surgery. We studied the effects of equi-MAC end-expiratory concentrations of isoflurane (0.65-0.75%) and nitrous oxide (60-65%). The anaesthetics were changed at random in three consecutive 10-min periods so that each patient received both drugs. From the AER recorded from the vertex and inion. Pa and Nb latency and amplitude were measured. N13, P20 latency and N13 amplitude were measured from SSER recordings from the neck and P15, N20. P25, N35, P45 latency and P15-N20, N20-P25, P25-N35 and N35-P45 amplitude from the scalp over the hand area of the sensory cortex. Compared with nitrous oxide, isoflurane significantly increased the latencies of the AER waves Pa (P = 0.02) andNb (P = 0.02), and the SSER waves N20 (P = 0.001) and P25 (P = 0.04). We were unable to demonstrate significant differences in Pa and Nb amplitude between isoflurane and nitrous oxide that we had seen previously. However, the amplitude of the SSER wave N20 was reduced significantly by nitrous oxide compared with isoflurane (P = 0.0004). This wave (N20) is thought to emanate from the thalamo-cortical radiations, and our findings may be explained by an analgesic effect of nitrous oxide mediated by endogenous opioids.

these various components could be assessed separately, then the development of an objective measure of depth of anaesthesia may be made easier. Separate assessment may also prove informative in the study of new anaesthetic drugs and may be useful in monitoring the additive or interactive effects of multiple drugs. Using evoked responses, we have attempted to separate the analgesic and hypnotic components of anaesthesia. Isoflurane is thought to be a strong hypnotic, in contrast with nitrous oxide, which is recognized as a weak hypnotic drug, but a good analgesic. General anaesthetics increase the latency and depress the amplitudes of the evoked responses [2]. In a previous study, we compared the effects of isoflurane and nitrous oxide on the auditory evoked response [3]. We found that, at concentrations of nitrous oxide and isoflurane that were equipotent, using the MAC concept [1] isoflurane had a more depressant effect on the auditory evoked response than nitrous oxide [3]—the latencies were increased and the amplitudes reduced. An opposite finding has been reported by others [4] for the somatosensory evoked response. Nitrous oxide had a more potent effect on the cortical somatosensory evoked responses than approximately equipotent concentrations of isoflurane. The changes in the auditory evoked response may reflect predominantly the hypnotic component of anaesthesia, while those in the somatosensory response may reflect mainly the analgesic component. In order to investigate this further, and to counter any arguments concerning inter-laboratory differences in anaesthetic estimations, we have extended our previous study [2] and recorded the auditory and the somatosensory evoked responses simultaneously.

DEPTH OF ANAESTHESIA: EVOKED RESPONSES

509

1 0 . 1 uV

Auditory

Somatosensory: scalp

Somatosensory: cervical

25

50

75

100

Latency (ms) FIG. 1. Auditory (AER) and somatosensory (SSER) evoked responses to show latency and amplitude measurements. Latency was measured from time zero to the peak or trough in question. For the AER, the amplitude was taken as the height of the vertical from the peak (in the case of a positive wave, e.g. Pa) to where it bisects the line joining two neighbouring troughs. For example, the dotted lines on the top trace indicate Pa and Nb amplitude. For the SSER, the amplitude was taken as a vertical from a peak (in the case of a positive wave) to the succeeding trough. For example, the dotted line on the bottom trace indicates N13-P20 amplitude.

surgery. All gave informed consent to the study, which was approved by the Hospital Ethics Committee. Each patient was premedicated with papaveretum 0.25 mg kg"1 and hyoscine 5 |ig kg"1 at least 1 h before induction of anaesthesia with thiopentone 2—4 mg kg"1. Vecuronium 0.1 mg kg"1 was given i.v. and the trachea intubated. Mechanical ventilation of the lungs was adjusted to maintain an end-tidal Pco2 of 4.0-5.5 kPa (Datex Capnomac). The inspired gas mixture was adjusted to produce the required endexpiratory concentrations: 60-65 % nitrous oxide or 0.65-0.75% isoflurane (Datex Capnomac); these concentrations were chosen to give approximately 0.6 MAC. The anaesthetic administered was changed randomly in each of the eight patients in three consecutive 10-min periods according to the treatment sequences recommended by Ebbutt [5]. The Capnomac analyser was calibrated using the manufacturer's calibration gas, and checked before and after each experiment using cylinders of isoflurane and Entonox, calibrated previously against volumetric standards using a gas chromatograph. Overpressure was used to reduce the time required to attain the target end-expiratory concentrations of isoflurane. Within 5 min, it was possible to have both stable end-expiratory values and residual concentrations of the previous agent less than 5 % of their original value. Arterial pressure and heart rate were recorded every 2.5 min throughout the procedure using a Datascope automatic recorder. Nasopharyngeal tem-

perature was recorded after tracheal intubation and at the end of the study period. Recording the evoked responses

The auditory stimulus was a rarefaction click presented to both ears simultaneously at 75 dB greater than the average hearing threshold at a stimulus repetition rate of 6 s"1. The somatosensory stimulus was a 150-(is electrical pulse produced by a constant current stimulator (Duostim) which had been modified by the addition of an optically isolated trigger input. This stimulus was applied to the median nerve at a rate of 2.2 s"1 and with sufficient current to produce a motor response. Both stimuli were applied continuously throughout the study. Purpose-built amplifiers were used to record three channels of EEG onto FM tape. Electrode placements for the auditory evoked response (AER) were Cz-In, for the cervical somatosensory evoked response (SSER) Pz-CVII and for the scalp SSER PzC3' (20 mm back from C3 on the 10-20 system [6]). The gain was 100 dB for all channels. On replay, signals were analog filtered with bandwidths of 25-500 Hz for the AER and 10-500 Hz for the SSER. The evoked responses were analysed using a digital signal processing card resident in an IBM PC and purpose-designed software. This program provided automatic rejection of artefact and a range of digital filtering. For these data we used a high pass filter of 20 Hz and three-point smoothing. The evoked responses were averaged over the last

Downloaded from http://bja.oxfordjournals.org/ at University of Ulster on June 25, 2015

P20

510

BRITISH JOURNAL OF ANAESTHESIA

Statistical methods Analysis of variance was used to compare the effects of the agents and the periods during which the agents were administered on the AER and SSER variables, and to determine if there was an interaction between the effect of the drug and the period in which it was given. This study design improves the power of the test for interaction, as it is based on within-subject variability rather than the between subject variability used in the standard two-period crossover design. RESULTS

Descriptive—AER The AER of a patient during an isoflurane-nitrous oxide—isoflurane randomization sequence is shown in figure 2. Isoflurane depressed the amplitude of the early cortical waves Pa and N b and increased their latencies, and when nitrous oxide was substituted this trend appeared to reverse. In this particular patient the amplitude and latencies in the second isoflurane period were not the same as those seen in the first isoflurane period. However, this effect was not statistically significant for the group as a whole. Descriptive—SSER Wave identification. The latencies of the waves of the individual SSER waveforms were measured. Average values (mean, SD) for the group of patients corresponded well with values reported previously. The prominent feature of the cervical SSER (fig. 1) was the negative wave N13 (14.5 (0.91) ms); Samra and colleagues [7] reported N13 as occurring at 14.8 ms at a similar anaesthetic concentration. This is followed by the positive wave which we have labelled P20 (20.4 (1.09) ms). The scalp SSER (figs 1, 3) was characterized by the positive wave which we have labelled P15 (16.3 (0.97) ms), followed by the negative N20 (20.6 (1.34) ms), and the positive P25 (25.7 (2.27) ms). McPherson and colleagues [4] have referred to these waves as the P1/N1/P2

0.5 nV

Isoflurane

Nitrous oxide

Isoflurane

25

75 100 50 Latency (ms) FIG. 2. Changes in the early cortical auditory evoked response of a patient who was given isoflurane, nitrous oxide and isoflurane for 10 min each. During the nitrous oxide period, Pa and Nb latencies decreased and their amplitudes increased compared with the isoflurane periods.

complex and the latencies they reported were 15, 20 and 23 ms, respectively. Following these, we found the negative wave N35 (36.5 (2.98) ms) and a positive wave occurring at 56.0 (5.75) ms, which we assumed to be P45 lengthened by the anaesthetic. Sometimes an additional peak (fig. 3) occurred around 25 ms. This was disregarded in making measurements. Measurement of the amplitudes and latencies of the various peaks in the individual patients yielded the following findings. In general, the cervical and scalp somatosensory latencies were increased with isoflurane compared with nitrous oxide. However, the latency of PI5 of the scalp SSER was relatively invariable between agents such that it was used as an aid to identification: the scalp SSER of a subject during a nitrous oxideisoflurane-nitrous oxide randomization is shown in figure 3. P15-N20 and N20-P25 amplitudes were greater and P25-N35 and N35-P50 amplitudes were less with isoflurane than with nitrous oxide. Statistical analysis There were no significant ( 5 % level) effects or interactions related to period for either the AER or SSER. The differences between the effects of nitrous oxide and isoflurane on the early cortical auditory evoked response are shown in table I. The results for the latencies are similar to those of the previous study [3] also shown in the table. Isoflurane increased the latencies with respect to nitrous oxide (the difference is therefore negative). The trends in the amplitudes were in the same direction in both studies; however, in this study the differences

Downloaded from http://bja.oxfordjournals.org/ at University of Ulster on June 25, 2015

2.8 min of each 10-min period when the endexpiratory concentrations were stable. Pa and N b latency and amplitude were measured using our previous system (fig. 1). Latency was measured from time zero to the peak or trough in question. Amplitude was measured as the height of the vertical from the peak (in the case of a positive wave, e.g. Pa) to where it bisects the line joining two neighbouring troughs. This has the advantage of counterbalancing fluctuations in the baseline caused by low frequency waves on which the AER waves are sometimes superimposed. In contrast with the AER, the SSER is plotted with the negative deflection upwards. This is in order to comply with a long established convention. For the SSER, N13, P20 latency and N13 amplitude were measured from the cervical recordings and P15, N20, P25, N35, P45 latency and P15-N20, N20-P25, P25-N35 and N35-45 amplitude from the scalp recordings. The measurement convention shown in figure 1 was used to quantify the SSER and is that quoted most commonly in the literature.

DEPTH OF ANAESTHESIA: EVOKED RESPONSES

0.5 uV

511

TABLE III. Somatosensory evoked response amplitude. Means for nitrous oxide (Nfi) and isoflurane (Iso.), differences between the means, standard error of the difference (SED) and significance of the difference

Means Nitrous oxide Cervically recorded N13-P20 Scalp recorded P15-N20 N20-P25 P25-N35 N35-P45

N,O

Iso. Difference

SED

3.6

3.9

-0.3

0.2

0.3

0.28 0.64 1.26 1.20

0.55 0.84 0.99 0.95

-0.28 -0.20 0.27 0.25

0.06 0.10 0.13 0.22

0.0004 0.07 0.08

P

0.3

Isoflurane 0.5 -

Iso. > N2O

Nitrous oxide

l

75

100

FIG. 3. Changes in the somatosensory evoked response recorded from scalp electrodes of a patient who was given nitrous oxide, isoflurane and nitrous oxide for 10 min each. During the isoflurane period, with the exception of PI 5 latency all the latencies increased compared with the nitrous oxide periods. Also P15-N20 and N20-P25 amplitudes increased and P25-N35 and N35-P45 reduced in the isoflurane period compared with the nitrous oxide period. TABLE I. Auditory evoked response latencies and amplitudes. Means for nitrous oxide (NjO) and isoflurane (Iso.), differences between the means, standard error of the difference (SED), and significance of the difference. Values from the previous study [3] are show in parentheses

Means

Latency (ms) Pa Nb

N,O

Iso.

35.0 52.6

40.7 62.5

Amplitude G»V) Pa 0.55 Nb 0.43

Difference

P

SED

-5.7 (-8.3) 2.4(2.5) 0.02(0.01) -9.9(-8.1) 4.0 (4.0) 0.02 (0.08)

0.37 0.18(0.54) 0.12(0.11) 0.16(< 0.001) 0.41 0.02 (0.38) 0.09(0.10) 0.8 (0.004)

TABLE II. Somatosensory evoked response latencies. Means for nitrous oxide (Nfi) and isoflurane (Iso.), differences betteeen the means, standard error of the difference (SED), and significance of the difference

Means N,O

Iso.

Difference

SED

P

14.4 20.2

14.6 20.8

-0.2 -0.6

0.1 0.3

0.07 0.06

16.4 16.4 20.2 21.2 24.8 26.7 35.9 37.3 54.1 58.0

0

-1.0 -1.9 -1.4 -3.9

0.1 0.2 0.8 0.7 2.4

0.001 0.04 0.09 0.16

Cervically recorded N13 P20

Scalp recorded P15 N20 P25 N35 P45

1.0

T3

0 - N15-P20

.P15-N20N20-P25

P25-N35 N35-P45

CD

"a E

-0.5 -I

N2O > Iso.

FIG. 4. The difference and standard error of the difference in the effects of nitrous oxide (N,O) and isoflurane (Iso.) on the somatosensory evoked response recorded from the neck and scalp. Positive values indicate that the attenuating effect of isoflurane is greater than that of nitrous oxide. * P = 0.0004.

between the two drugs were much smaller and not significantly different from zero. The differences between the effects of nitrous. oxide and isoflurane on the SSER latencies are shown in table II. Significant differences occurred in N20 and P25. The differences between the effects of nitrous oxide and isoflurane on the SSER amplitudes are shown in table III and plotted in figure 4. The amplitudes in the early section of the response (N13-P20, P15-N20 and N20-P25) were greater with isoflurane than with nitrous oxide and in the later section (P25-N35 and N35-P45) were greater with nitrous oxide than with isoflurane. A significant difference occurred in PI5—N20. Other physiological variables

Systolic arterial pressure and heart rate were significantly greater in the first 10-min period, which included tracheal intubation, compared with the second and third periods (systolic pressure difference 5.8 (SED 1.1) mm Hg) (P = 0.0002); heart rate difference 18.4 (2.6) beat min"1 (P = 0.0001), as would be

Downloaded from http://bja.oxfordjournals.org/ at University of Ulster on June 25, 2015

I 50 Latency (ms)

25

512

BRITISH JOURNAL OF ANAESTHESIA

expected. Systolic arterial pressure was slightly but significantly less (4.6 (1.1) mm Hg) with administration of isoflurane compared with nitrous oxide (P < 0.001). In all patients, temperatures remained within the range 35.5—36.7 °C during the period of investigation and there were no effects that could be attributed to differences between the periods or drugs. The patients appeared to be anaesthetized adequately during the study, and no patient volunteered recall of any event during anaesthesia. DISCUSSION

Downloaded from http://bja.oxfordjournals.org/ at University of Ulster on June 25, 2015

The hypothesis tested in the study was that the early cortical auditory evoked response represented by waves Pa and Nb reflects the hypnotic component of anaesthesia and would be expected to be affected more by isoflurane than by nitrous oxide. In contrast, the somatosensory evoked response, in particular wave N20, reflects the analgesic component of anaesthesia, and would be affected more by nitrous oxide than by isoflurane. In constructing this hypothesis, isoflurane was assumed to be a potent hypnotic and weak analgesic and nitrous oxide a weak hypnotic and potent analgesic. This requires qualification. First, we need to define "hypnosis" and "analgesia". Prys-Roberts [8], in an attempt to define these terms, started with the premise that pain is the conscious perception of a noxious stimulus. He defined anaesthesia as that state in which, as a result of a drug-induced unconsciousness, the patient neither perceives nor recalls noxious stimulation. Hypnosis was conceived as synonymous with anaesthesia, as it implies a drug-induced sleep and analgesia as diminished or abolished perception of pain in an otherwise conscious patient. However, we feel that the concept of analgesia may be extended to the anaesthetized patient. For example, analgesic drugs such as fentanyl and morphine reduce the amount of inhalation agent required to anaesthetize a patient. Using the concept of MAC (minimum alveolar concentration which prevents movement in response to surgical incision in 50 % of patients) as a measure of anaesthetic potency [1], morphine reduces MAC for the inhalation agents by up to 75 %. However, neither fentanyl nor morphine alone produces anaesthesia. It is clear from this that there are at least two (there may be more) components of anaesthesia, an analgesic one provided by drugs such as fentanyl and morphine and an hypnotic one, which "puts the patient to sleep". This study supports our hypothesis that there are aspects of the SSER which reflect analgesia rather than hypnosis, in that we have demonstrated a greater reduction in amplitude of PI5—N20 with nitrous oxide compared with isoflurane. This is in contrast with our findings in the AER, where there was no significant difference between the effects of each anaesthetic on Pa and Nb amplitude. In the present study, there were differences in the AER results compared with our previous study, but these did not affect our conclusions. In this study, the changes in the early cortical waves Pa and Nb were not affected by period. In our previous study Pa and

Nb had shown greater amplitudes and shorter latencies in the first period compared with the second and third. Also, the reduction in Pa and Nb amplitudes by isoflurane compared with nitrous oxide were not significant in this study, whereas previously they were. One possible explanation for this is that the simultaneous somatosensory stimulation may have lightened the depth of anaesthesia under isoflurane, and hence decreased the difference in effect between the two agents. Clearly, it is necessary to study more subjects in order to resolve this discrepancy. The studies most comparable to ours on the relative effects of isoflurane and nitrous oxide on the median nerve somatosensory evoked response are those of McPherson and colleagues [4], Restuccia and colleagues [9] and Peterson, Drummond and Todd [10]. McPherson's group found, as we have, that isoflurane increased N20 latency more than nitrous oxide, and that nitrous oxide reduced N20 (N20-P25 in our study) amplitude more than isoflurane. The isoflurane concentrations in their study were 0.25-1.0% (0.22-0.87 MAC) and this was compared with 60% nitrous oxide (0.48 MAC). Restuccia and colleagues [9] found that N20 latency was greater and the amplitude (PI5—N20) was depressed less with 1.0 MAC isoflurane on its own than with 0.5 MAC isoflurane and 67 % nitrous oxide in oxygen (total of 1.14 MAC). Our study supports these findings. Peterson, Drummond and Todd came to different conclusions. They found that N20 latency was greater, but that the amplitude (N20-P25) was depressed more by 1.5 MAC isoflurane on its own than by a 1.58 MAC mixture of isoflurane and nitrous oxide. At 1.5 MAC isoflurane, N20 could be detected in only four of the seven patients, compared with seven of seven when they were receiving the 1.58 MAC mixture. The suggestion here is that 0.5 MAC isoflurane reduced the amplitude of N20 (N20-P25) more than 0.58 MAC nitrous oxide. Other studies report practically no effect of nitrous oxide on N20 latency [11, 12], whereas a definite and substantial increase has been reported with isoflurane [7, 13]. Both nitrous oxide and isoflurane are reported to reduce the amplitude of N20 [7, 11-13], but precisely which agent has the greater effect cannot be determined from these studies. In addition to comparing the relative effects of nitrous oxide and isoflurane on the AER and SSER, this investigation allowed us to compare relative effects of anaesthetics in different sections of the somatosensory neuraxis. Others have found that the earlier waves were relatively robust to the effects of anaesthetics [7, 9, 12]. Not surprisingly, therefore, we found no significant differences between the two drugs in the effects on the cervically recorded waves and the scalp recorded wave PI5. These waves are thought to be generated at the level of, or caudal to the medial lemniscus [14]. Nitrous oxide therefore, does not appear to antagonize nociceptive transmission at that level. N20 wave latency was the earliest to show a statistically significant difference. The subsequent wave P25 also showed a statistically significant

DEPTH OF ANAESTHESIA: EVOKED RESPONSES

mechanism has been reviewed extensively by Finck [21] and rests on the antagonism of the analgesic effects by naloxone, and the development of tolerance. Second, nitrous oxide may act directly on the cortex to diminish pain perception, which accords with our own findings [22] and those of others [23, 24] that nitrous oxide depresses that part of the auditory evoked response thought to be generated by the primary and frontal cortex and association areas. In conclusion, we have demonstrated a clear difference in the effect of nitrous oxide on the N20 wave of the SSER when compared with isoflurane. This contrasts with cortically derived waves of both the AER and the SSER, on which the effects of the two drugs were similar. These changes in the evoked responses could provide a basis for further research into the effects of both hypnotic and analgesic drugs used in anaesthesia. REFERENCES 1. Eger El n. Anaesthetic Uptake and Action, 5th Edn. Baltimore: Williams and Wilkins, 1974. 2. Thornton C. Evoked potentials in anaesthesia. European Journal of Anaesthesiology 1991; 8: 89-107. 3. Newton DEF, Thornton C, Creagh-Barry P, Dore CJ. The early cortical auditory evoked response in anaesthesia: comparison of the effects of nitrous oxide and isoflurane. British Journal of Anaesthesia 1989; 62: 61-65. 4. 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: 626-633. 5. Ebbutt AF. Three-period crossover designs for two treatments. Biometrics 1984; 40: 219-224. 6. Jasper HH. Report on the committee on methods of clinical examination in electroencephalography. Electroencephalography and Clinical Neurophysiology 1958; 10: 370-375. 7. Samra KS, Vanderzant CW, Domer PA, Sackellares JC. Differential effects of isoflurane on human median nerve somatosensory evoked potentials. Anesthesiology 1987; 66: 29-35. 8. Prys-Roberts C. Anaesthesia: A practical or impractical construct? British Journal of Anaesthesia 1987; 59: 1341-1345. 9. Restuccia D, Primieri P, Di Lazzaro V, Lo Monaco M, Bonomo V, Gualtieri E, Villani A. Effets du protoxyde d'azote sur les potentiels evoques enregistres pendant PanestMsie a l'enflurane et a Pisoflurane [The effects of nitrous oxide on evoked potentials recorded during enflurane and isoflurane anaesthesia]. Cahiers a"Anesthisiologie 1990; 38:91-94. 10. 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. 11. Sloan TB, Koht A. Depression of cortical somatosensory evoked potentials by nitrous oxide. British Journal of Anaesthesia 1985; 57: 849-852. 12. Schubert A, LicinaMG, LineberryPJ. The effect of ketamine on human somatosensory evoked potentials and its modification by nitrous oxide. Anesthesiology 1990; 72: 33-39. 13. Nogueira MC, Brunko E, Vandsteen A, DeRood M, Zegers de Beyl D. Differential effects of isoflurane on SER recorded over parietal and frontal scalp. Neurology 1989; 39: 1210-1215. 14. Chiappa KH. Short latency somatosensory evoked responses: interpretation. In: Chiappa KH, ed. Evoked Potentials in Clinical Medicine. New York: Raven Press, 1985; 251-324. 15. Abbruzzesse M, Favale E, Leandri M, Ratto S. New subcortical components of the cerebral somatosensory evoked potential in man. Ada Neurologica Scandinavica 1978; 58: 325-332.

Downloaded from http://bja.oxfordjournals.org/ at University of Ulster on June 25, 2015

difference. In all cases where there was a change, the latencies were longer with isoflurane than nitrous oxide. It would seem, therefore, that isoflurane slows transmission more than nitrous oxide from around 13-20 ms onwards and, because more synapses have to be crossed to reach to the higher regions of the tract, the waves originating from higher centres are more affected than those generated more caudally. Also, it appears that changes in wave latency reflect hypnosis rather than analgesia. N20 amplitude, when measured from the preceding peak, Pi5, showed a statistically significant difference; when measured to the succeeding wave, P25, which is the amplitude measurement reported most often in the literature, the difference between the two drugs approached significance (P = 0.07). In both these cases, nitrous oxide reduced the amplitude compared with isoflurane. In contrast, after 25 ms the amplitudes of P25—N35 and N35—P45 were greater for nitrous oxide than for isoflurane, but these differences were not statistically significant, although for P25-N35 amplitude it was close (P = 0.08). N20 and P25 originate from the region of the thalamus and primary cortex [14-16]. The amplitude of N20 (latency span of 15-25 ms) was depressed more by nitrous oxide than by isoflurane and we speculate that, between the medial lemniscus and primary somatosensory cortex, there are structures involved in analgesia. Rostral to the primary somatosensory cortex are the frontal cortex and association areas, represented by waves N35 and P45. These structures may be involved more in hypnosis, in that the amplitudes measured in this study were similar and both agents were probably exerting a similar depressive effect. Two sources of information support these ideas. Freye, Buhl and Ciaramelli [17] found that N20 decreases in a dose-related way with alfentanil, an opioid binding to the u receptor. This can be antagonized by increasing the stimulus amplitude or by administering naloxone. Alfentanil had no effect on a later N100 wave which they investigated. The likely generators of this wave, which we did not study, are also in the frontal cortex and association areas. In contrast, nalbuphine, an opioid with K agonist activity, did not affect N20, but reduced the amplitudes of N100. This was antagonized by increasing the stimulus amplitude, but not by naloxone. The authors suggested that, as N20 is primarily generated in the pontine-thalamic region, die mode of action of fentanyl resembles that of the block of sensory impulses. This takes place before impulses are transmitted to more rostral structures responsible for the identification of pain. The effect of nalbuphine on late evoked potentials may indicate an activity on the thalamo-cortical projection sites which are involved in perception of pain. This could be the mechanism for an analgesic effect attributed, to a greater or lesser extent, to all volatile anaesthetic agents when they are used in small concentrations for obstetric analgesia [18-20]. Nitrous oxide may, therefore, exert its analgesic effect in two ways. First, it may stimulate the production of endogenous opioids which act at receptors in the pain pathways. (This would explain the strong attenuation of N20 emanating from the thalamus.) The evidence for this

513

514 16. Goldie WD, Chiappa KH, Young RR, Brooks EB. Brainstcm auditory and short latency somatosensory evoked responses in brain death. Neurology 1981; 31: 248-256. 17. Freye E, Buhl R, Ciaramelli F. Opioids with different affinity for subreceptors induce different effects on early and late sensory evoked potentials (SEP) in man. National Institute for Drug Abuse Research Monograph Service 1986; 75: 551-554. 18. Rosen M, Mushin WW, Jones PL, Jones EV. Field trial of methoxyflurane, nitrous oxide and trichlorethylcne as obstetric analgesics. British Medical Journal 1969; 3: 263-267. 19. Abboud TK, Shnider SM, Wright R. Enflurane analgesia in obstetrics. Anesthesia and Analgesia 1981; 60: 133-137. 20. Abboud TK, Gangolly J, Mosaad P, Crowell D. Isoflurane in obstetrics. Anesthesia and Analgesia 1989; 68: 388-391.

BRITISH JOURNAL OF ANAESTHESIA 21. Finck D. Nitrous oxide analgesia. In: Eger El II, ed. Nitrous Oxide/N,O. London: Edward Arnold, 1985; 41-55. 22. Thornton C, Barrowcliffe MP, Konieczko KM, Vcntham P, Dore CJ, Newton DEF, Jones JG. The auditory evoked response as an indicator of awareness. British Journal of Anaesthesia 1989; 63: 113-115. 23. Lader MH, Norris H. Effect of nitrous oxide on the auditory evoked response in man. Nature {London) 1968; 218: 1081-1082. 24. Harkins SW, Benderti C, Colpitts YH, Chapman CR. Effects of nitrous oxide inhalation on brain potentials evoked by auditory and noxious dental stimulation. Progress in NeurcPsychopharmacology and Biological Psychiatry

1982; 6:

164-174.

Downloaded from http://bja.oxfordjournals.org/ at University of Ulster on June 25, 2015