Full text of DOI:10.1007/BF03018938

494
Reports of Investigation
Median nerve evoked
responses and explicit
memory during recov-
ery from isoflurane/
nitrous oxide anesthesia
Ingrid Rundshagen MD,
Kai Schnabel,*
Jochen Schulte am Esch MD,†
Purpose: To evaluate median nerve somatosensory evoked responses during recovery from anesthesia in relation to clin-
ical findings.
Methods: Twenty-two gynecologic patients received isoflurane in nitrous oxide for anesthesia. Midlatency somatosenso-
ry evoked responses (N20, P25, N35, P45, N50) were recorded the day before surgery (AWAKE), during steady state
anesthesia (STABLE), and every five minutes after discontinuation of anesthesia until the patients were able to name a
shown object correctly (RECOVERY). Next day the patients were questioned with a structured interview about their
explicit memory of the immediate recovery period and classified into groups: No-MEM (no memory) and MEM (memo-
ry). Multivariate analysis of variance compared electrophysiological parameters at the different time points and between
the two memory groups.
Results: During STABLE isoflurane/N2O anesthesia, all cortical amplitudes were reduced (P # 0.003) and all latencies
were prolonged compared with AWAKE (P < 0.001). At RECOVERY the latencies N35, P45, N50 remained prolonged
(P # 0.001), while the amplitudes N20P25 and P45N50 were reduced in comparison to AWAKE (P # 0.02). The laten-
cies P45 (48 ± 8 vs 61 ± 9 msec) and N50 (67 ± 12 vs 81 ± 10 msec) were shorter in the patients of the group MEM
(P # 0.03) at RECOVERY.
Conclusion: The reversibility of anesthetic induced changes in amplitudes and latencies of median nerve somatosensory
evoked responses reflected clinical awakening during emergence from isoflurane/nitrous oxide anesthesia. In the patients
who had recall for the immediate recovery period, the reversibility of anesthetic induced changes of components P45 and
N50 was faster than in patients without recall.
Objectif : Évaluer les potentiels évoqués somesthésiques du nerf médian en relation avec les constatations cliniques pen-
dant la récupération de l’anesthésie.
Méthode : Vingt-deux patientes de gynécologie ont reçu une anesthésie avec un mélange d’isoflurane et de protoxyde d’a-
zote. Les potentiels évoqués somesthésiques de milatence (N20, P25, N35, P45, N50) ont été enregistrés le jour précédant
l’opération (VIGILE), pendant une période stable de l’anesthésie (STABLE) et toutes les cinq minutes après l’arrêt de
l’anesthésie jusqu’à ce que les patientes puissent nommer correctement un objet désigné (RÉCUPÉRATION). Le jour sui-
vant, on a interrogé les patientes lors d’une rencontre structurée au sujet de leur mémoire explicite de la période de récupéra-
tion immédiate. On les a ensuite réparties en groupes : No-MEM (sans mémoire) et MEM (mémoire). L’analyse de variance
multivariée a permis de comparer les paramètres électrophysiologiques entre les deux groupes à différents moments.
Résulltats : Pendant l’action STABLE de l’anesthésie à l’isoflurane/N2O, toutes les amplitudes corticales ont été réduites (P
# 0,003) et tous les temps de latence ont été prolongés, comparés aux données VIGILES (P < 0,001). Pendant la
RÉCUPÉRATION, les temps de latence N35, P45, N50 sont demeurés longs (P # 0,001), tandis que les amplitudes
N20P25 et P45N50 ont été réduites, comparées aux amplitudes VIGILES (P # 0,02). Les temps de latence P45 (48 ±
8vs 61 ± 9 msec) et N50 (67 ± 12 vs 81 ± 10 msec) ont été plus courts chez les patientes du groupe MEM (P # 0,03)
à la RÉCUPÉRATION.
Conclusion : La réversibilité des changements, induits par l’anesthésique, d’amplitude et de temps de latence des poten-
tiels évoqués somesthésiques du nerf médian correspondait au réveil clinique pendant la récupération de l’anesthésie à
l’isoflurane/N2O. Chez les patientes qui ont eu une mémoire immédiate de la récupération, la réversibilité des change-
ments provoqués par l’anesthésique des composantes P45 et N50 a été plus rapide que chez les patientes sans souvenir.
From the University Hospital Charité, Department of Anesthesiology, Berlin, Max-Planck-Institute for Human Development,* Berlin and
Department of Anesthesiology, University Hospital Eppendorf,† Hamburg, Germany.
Address correspondence to: Ingrid Rundshagen MD, Department of Anesthesiology and Intensive Care, University Hospital Charité,
Campus Charité Mitte, Schumannstr. 20/21, D-10117 Berlin, Germany. Phone: 49-30-2802-2808; Fax: 49-30-2802-5065; E-mail:
ingrid.rundshagen@charite.de
Parts of this work were presented at the 24th annual meeting of the Society of Neurosurgical Anesthesia and Critical Care (SNACC), New
Orleans, USA, 1996.
Accepted for publication February 11, 2000.
CAN J ANESTH 2000 / 47: 6 / pp 494–502

Rundshagen et al.: MnSSER DURING RECOVERY
495
SSESSING depth of anesthesia is a major
concern of anesthesiologists in avoiding
intraoperative awareness. The introduction
be reversed in relation to clinical awakening.
Moreover, we hypothesised that, if the patients differ
in explicit memory performance the reversibility of
MnSSER changes would be delayed in the patients
without recall for the recovery period.
A
of balanced anesthetic techniques involving
neuromuscular blocking agents makes the assessment
more difficult because respiratory patterns and spon-
taneous movements as signs of inadequate anesthesia
are no longer available. Alternative approaches that
have been proposed include a more direct measure-
ment of central nervous activity using the spontaneous
or evoked electroencephalogram.
Methods
Subjects
After approval by the ethical committee and written
informed consent, 22 female patients (age: 39 10 yr,
weight: 65 14 kg, height: 167 10 cm, ASA classi-
fication I – II) scheduled for breast or lower abdomi-
nal gynecological surgery were included in the study.
Patients with neurological diseases or who took CNS-
active drugs were excluded.
Somatosensory evoked responses are used clinically
to monitor neural structures at risk during neurosur-
gical, orthopedic, or vascular surgery. They can be
obtained by repetitive presentation of a suitable stim-
ulus to the pathway at risk.1 There are three negative
and two positive midlatency MnSSER wave compo-
nents in the awake state.2 The first component N20
probably represents the synaptic events triggered by
the primary input volley at the contra-lateral cortical
projection area of the hand, and generators are
assumed to originate in the area 3 b of the primary
sensory cortex (Brodman’s classification).3 Generators
for the later deflections are less well-defined and may
originate from both cortico-cortical connections and
separate thalamo-cortical projections.4,5 Using brain
electrical source analyses Srisa-an et al. demonstrated
that N60 is most likely generated in area 1.6 However,
despite several neurophysiological investigations using
either electrocorticography or, more recently, results
from magnetic responses to electrical median nerve
stimulation in combination with imaging techniques,
the different generators of the SEP components are
still debated.7–9
As a possible application for monitoring depth of
anesthesia, median nerve somatosensory evoked
responses (MnSSER) have frequently been studied
during induction and maintenance of anesthesia.10,11
Thornton et al. suggested that the effect of nitrous
oxide on the subcortical waves reflected analgesia,
while the later cortical waves were sensitive to the hyp-
notic potency of isoflurane.1 2 In contrast, during
emergence from anesthesia, MnSSER have rarely been
recorded. Moreover, no study investigated MnSSER
components in relation to memory function during
anesthesia. However, midlatency auditory evoked
potentials provide information on the possibility of
perceptual processing during general anesthesia.1 3
This prospective study investigated the relationship
between MnSSER and clinical signs of awakening dur-
ing recovery from isoflurane/nitrous oxide anesthesia.
We hypothesised that the anesthetic induced changes
of midlatency MnSSER components would gradually
Anesthesia
The patients received pre-medication with 7.5 mg
midazolam po 45 min before induction of anesthesia
with 0.3 mg·kg–1 etomidate and 1.5 µg·kg– 1 fentanyl
followed by 0.1 mg·kg–1 vecuronium. After tracheal
intubation, anesthesia was maintained with end-tidal
isoflurane 0.6 Vol% in nitrous oxide (66%) in oxygen
(Cicero EM, F. Dräger, Lübeck, Germany). If neces-
sary, additional bolus injections of 0.5 µg·kg–1 fentanyl
were administered after 30-45 min according to the
surgical procedure, if the mean arterial blood pressure
and/or heart rate increased more than 20% compared
to baseline values. After surgery, when the MnSSER-
recordings had been performed, isoflurane and
nitrous oxide were discontinued, and the lungs were
ventilated with oxygen 100%. Controlled ventilation
was continued as long as the patients tolerated the tra-
cheal tube. When the patients started coughing or
showed spontaneous movements and responded to a
verbal command to take a deep breath, extubation was
performed. Subsequently, oxygen was supplied via a
facial mask. The MnSSER recording was continued
until the patients opened their eyes spontaneously.
MnSSER aquisition
On the day before surgery, patients became accustomed
to the experimental setting and comfortably relaxed with
closed eyes in a supine position for MnSSER-recording
(Evomatic 4000® system [Dantec, Copenhagen,
Denmark]). The baseline MnSSER recordings were per-
formed after the assessment of the individual stimulation
thresholds. For unilateral stimulation of the median
nerve (stimulation rate 3 Hz), a mounted dual stimula-
tor (13L36, Dantec, Copenhagen, Denmark) with the
cathode 2.3 cm from the anode produced monophasic
rectangular pulses of 0.2 msec duration at the wrist. By
preference the right median nerve was stimulated.

496
CANADIAN JOURNAL OF ANESTHESIA
However, in five patients for whom dissection of axillary
lymph nodes on the right side was planned for breast
cancer, the left median nerve was stimulated. In relation
to the individual sensory and motor thresholds first the
threefold sensory threshold intensity and then the motor
plus sensory threshold intensity was used as dummy
blocks. Then the stimulation intensity was further
increased to the individual threshold of tolerance and
two baseline recordings were performed. The stimula-
tion intensity was kept constant throughout the experi-
ment. The MnSSER waveforms were recorded
simultaneously on three amplifier channels using sterile
platinum needle electrodes (13L70, Dantec,
Copenhagen, Denmark) placed over the ipsilateral
brachial plexus (Erb´s point), the spinous process of the
6th cervical vertebra (C6) and the contralateral cortical
hand area at the scalp (C3´ or C4´).2 A mid-forehead ref-
erence electrode was used (Fpz; international 10-20
System) and electrode impedances were kept below 10
kS. The bandpass was set at 20 Hz to 2 kHz.1 4 a The
device includes an automatic artefact rejection mode. For
data acquisition, the digital resolution was 12 bits, the
sampling rate was 25 kHz per channel, and the resolu-
tion 1000 points per channel. A post-stimulus period of
90 msec (180 msec during recovery from anesthesia) was
analysed. For each response, 200 stimuli were averaged
and stored on disk for later analysis. Latencies and ampli-
tudes of MnSSER were obtained using a software pack-
age (EvoPC®, Müller, Hamburg, Germany). The
following peak latencies were evaluated: at Erb N10; at
C6 N13 and P17; at the scalp a w-shaped complex with
the three prominent negative (N20, N35, N50) and two
prominent positive components in between (P25, P45)
occurring within 15 to 70 msec post stimulus after medi-
an nerve stimulation.2 During anesthesia and recovery,
the first prominent negative peak (occurring between
17-25 msec) was defined as N20 and the following com-
ponents were subsequently examined.2 The next positive
wave was referred to as P25, the second negative peak
was referred to N35, etc. Furthermore, the peak-to-peak
amplitudes N13P17, N20P25, P25N35, N35P45,
P45N50 and the central conduction time (CCT; latency
N20 – latency N13) were calculated.
recording before extubation and Post-EXT was defined
as the first recording following extubation, when the
patients opened their eyes on command. When the
patients opened their eyes spontaneously, they were
asked to define verbally a shown object (a small red
booklet [about 12 × 15 cm in size], which was opened
and closed in front of the patient). When they correct-
ly named the shown object, they were asked to keep it
in mind. Then, the final MnSSER recording was per-
formed (RECOVERY).
Vital parameters and clinical findings
The electrocardiogram (HR; beats per minute), mean
arterial blood pressure (MAP; mmHg), pulse oximet-
ric oxygen saturation (SpO ;%), capnography for CO
(PETCO2, mmHg), end exs2piratory isoflurane concen2-
tration (IsoET, Vol %) and nitrous oxide (Vol %) con-
centrations during and after anesthesia (after
extubation a tight face mask was used and the patients
were instructed breathe in deeply), skin temperature
of the stimulated arm and rectal temperature (C) were
recorded simultaneously (Cicero EM, F. Dräger,
Lübeck, Germany). The total time of anesthesia, the
time of the extubation and the time when the patients
were able to name the shown object were registered.
Memory
The day after the operation, the patients were asked,
using a structured interview, about the immediate
recovery period (Table I). Based on their answers they
were classified into two groups: No-MEM - without
memory; MEM - with explicit memory for the imme-
diate recovery period.
Statistical analysis
Means and standard deviations of values for the vari-
ous peak latencies, conduction times, and amplitudes
were calculated. The distribution of data was tested
with a Kolmogorov-Smirnov test. Numerical data
regarding MnSSER components and vital parameters
were subjected to repeated measures analysis of vari-
ance using Hotellings T-square.1 5 This statistical
The MnSSER recording was performed the day
before surgery (AWAKE), during steady state isoflu-
rane/nitrous oxide anesthesia after the end of surgery
(STABLE), and every five to ten minutes during recov-
ery from anesthesia. In accordance with the clinical
findings, Pre-EXT was defined as the last MnSSER-
TABLE I
Structured interview 24 hr after anesthesia:
1
2
3
4
What was the first you remembered after waking up from
anesthesia?
Do you remember anything shown to you during the wake-up
period?
Do you remember whether the MnSSER recording was
performed?
Were you asked to keep anything in mind during recovery
from anesthesia?
a For further studies a high pass setting of 1-5 Hz is suggested
(s. discussion)

Rundshagen et al.: MnSSER DURING RECOVERY
497
TABLE II Vital parameters
AWAKE
STABLE
Pre-EXT (9 5 min)
No-MEM / MEM
Post-EXT (15 5 min)
No-MEM / MEM
RECOVERY (22 5 min)
No-MEM / MEM
No-MEM / MEM
No-MEM / MEM
HR /min
MAP (mmHg)
SpO2 (%)
71 2 / 70
90 3 / 84
98 1 / 99
2
4
1
67 3 / 70
80 5 / 78
99 1 / 99
34 1 / 35
4
4
1
1
73 5 / 75
98 5 / 96
99 1 / 99
37 2 / 38
6
6
1
1
89 4 / 89
102 5/ 96
99 1 / 99
39 1 / 37
6
89 4 / 94
92 4 / 91
99 1 / 99
39 1 / 38
6
4
1
2
5
1
1
PCO (mmHg)
2
IsoET (%)
0.6 0.1 / 0.6 0.1
61 4 / 60
36.4 0.3 / 36.7 0.4 35.3 0.3 / 35.7 0.2
0.2 0.1 / 0.2 0.1
17 6 / 13
35.3 0.4 / 35.7 0.2 35.5 0.4 / 35.9 0.2 35.5 0.4 / 35.9 0.2
0.1 0.1 / 0.1 0.1
0.1 0.1 / 0.1 0.1
N2OET
(
%)
5
4
7
2 / 8
2
4
1 / 5
2
Rectal TEMP (°C)
Means SD of heart rate (HR), mean arterial blood pressure (MAP), pulse oxymetric oxygen saturation (SpO ), end exspiratory carbon
2
dioxide (PCO), isoflurane (IsoE T) and NO (NOE T) concentrations and rectal body temperature (TEMP) before, during steady state
2
2
2
isoflurane/nitrous oxide anesthesia and during emergence from anesthesia. No-MEM = patients without memory of the wake-up phase,
MEM = patients with memory. Measurements were performed before (AWAKE), during isoflurane/nitrous oxide anesthesia (STABLE)
and during recovery from anesthesia (Pre-EXTubation; Post-EXTubation, RECOVERY).
Results
TABLE III MnSSER latency changes in relation to memory
function
Vital parameters and clinical findings
The mean time of anesthesia was 96 36 min. The
patients received 0.2 0.05 mg fentanyl in addition to
the induction dose with a mean interval of 61 18
min before the end of anesthesia. Tracheal extubation
was performed 12 5 min after the end of anesthesia.
At 22 5 min after discontinuation of anesthesia the
patients were able to name the shown object correct-
ly. During the wake-up phase, the hemodynamic para-
meters were stable. There was no request for any
additional medication. At RECOVERY the patients
reported no pain.
AWAKE
RECOVERY
P45
No-MEM
MEM
No-MEM
MEM
41.5 7.3
39.4 4.6
55.4 6.5
52.0 5.9
60.6 7.8*
47.8 8.1
81.1 9.8*
67.3 12.1
N50
Means and SD of MnSSER latencies P45 and N50 before
(AWAKE) and 22 5 min after the end of isoflurane/nitrous
oxide anesthesia (RECOVERY). The group No-MEM includes
patients without explicit memory for the immediate wakeup phase
(n = 9, evaluated 24 hr later), MEM includes patients who pre-
sented recall for the recovery period (n = 10). * = P < 0.05
between the two groups.
The day after the operation, 19 patients could be
interviewed about their memory for the recovery peri-
od, because three had already been discharged from
the hospital. While nine patients did not remember
the wake-up phase, 10 showed explicit memory. Seven
patients remembered the shown book, and three
recalled the electrical stimulation of the median nerve
at their wrist. Only two of the seven patients who
remembered the shown object recalled the nerve stim-
ulation as well. There were no differences in demo-
graphics, fentanyl consumption or delivery, the time
of the extubation, the time to identify the shown
object correctly after anesthesia, the vital parameters
or the end-tidal gas concentrations between the two
groups (Table II). However, the time of anesthesia
was longer in the group No-MEM than in the group
MEM (110 26 min vs 82 31 min; P < 0.05).
approach is most appropriate for studies involving
repeated measurements in the same individual at dif-
ferent times, but has the disadvantage of being applic-
able only to complete data sets for all time points. For
descriptive purposes, differences between different
time points were analysed with paired t tests for all
measurements.
Between group comparisons (MEM vs No-MEM)
regarding the patients’ characteristics and clinical find-
ings were conducted by Students’ t tests. Statistical
comparisons of all MnSSER components were per-
formed with multivariate analyses (Hotellings T-
square; independent variable: memory) at AWAKE
and RECOVERY between the memory groups. The
multivariate comparisons at STABLE were limited to
components N20, P25 and N35 due to missing data,
when the components of the later MnSSER compo-
nents were suppressed during anesthesia. P < 0.05 was
adopted for level of significance for all statistical tests.
MnSSER components
The mean stimulation intensity was 12 3 mA, while the
motor threshold reached 6 3 mA. All cortical MnSSER
components were well defined preoperatively except
P45, which was absent in one patient, and N50, which

498
CANADIAN JOURNAL OF ANESTHESIA
FIGURE 1 Means SD of latencies of MnSSER components
before (AWAKE), during steady state isoflurane/nitrous oxide
anesthesia (STABLE) and during recovery from anesthesia (Pre-
EXT = before extubation, Post-EXT = after extubation, RECOV-
ERY = 22 5 min after discontinuation of isoflurane/nitrous
oxide). N = 22 patients, * = P < 0.05 vs AWAKE.
FIGURE 3 MnSSER latency components P45 and N50 before
(AWAKE) and 22 5 min after discontinuation of
isoflurane/nitrous oxide (RECOVERY). Group No-MEM (n = 9)
includes patients without memory for the immediate recovery
period, while group MEM (n = 10) includes patients with explicit
memory function evaluated 24 hr later.
FIGURE 4 Midlatency MnSSER components (90 msec post-
stimulus interval, C3‘ vs Fpz) due to electrical stimuli delivered to
the right median nerve at the wrist before (AWAKE), during
steady state isoflurane/nitrous oxide anesthesia (STABLE) and
during recovery from anesthesia (Pre-EXT = before extubation,
Post-EXT = after extubation, RECOVERY = 25 min after discon-
tinuation of isoflurane/nitrous oxide) in a female subject with
recall for the wakeup phase, 32 yr, ASA I.
FIGURE 2 Means SD of amplitudes of MnSSER components
before (AWAKE), during steady state isoflurane/nitrous oxide
anesthesia (STABLE) and during recovery from anesthesia (Pre-
EXT = before extubation, Post-EXT = after extubation, RECOV-
ERY = 22 5 min after discontinuation of isoflurane/nitrous
oxide). N = 22 patients, * = P < 0.05 vs AWAKE.
was absent in two patients. Satisfactory traces were
obtained in all patients during steady state anesthesia and
the wakeup phase. The Kolmogorov-Smirnov test sup-
ported the assumption of normal distribution for all
MnSSER parameters. The multivariate analyses revealed
differences for CCT and all cortical latencies, comparing
the awake values with the components during anesthesia
of the different peaks. An interaction between latency
components and time revealed a more pronounced effect
on the later MnSSER waves (P = 0.0002). The impor-
tance of this effect is remarkable regarding the higher
between-subject variance in those components which
tends to mask effects in the inferential statistical analysis.
Mean values of cortical peak-to-peak midlatency ampli-
tudes are shown in Figure 2.
and the measurements during the wakeup phase (
P <
0.004). Figure 1 summarises the results for the latencies

Rundshagen et al.: MnSSER DURING RECOVERY
499
During STABLE the primary cortical complex
N20P25 remained identifiable in all patients. In contrast,
at STABLE, N35 was identified in only 18 patients, P45
in 13 and N50 in 11 patients. The latencies of the corti-
cal components and CCT were significantly prolonged
Discussion
In the present study we investigated anesthetic induced
changes of midlatency MnSSER components during
steady state and recovery from isoflurane/nitrous oxide
anesthesia in relation to clinical findings and awake
baseline values. We demonstrated during 1 MAC isoflu-
rane/nitrous oxide anesthesia, that the anesthetic
effects increased with the latency of the cortical
MnSSER components, which is in agreement with pre-
vious studies including non-specific later MnSSER
components.1 6 When, during recovery from anesthesia,
the patients opened their eyes spontaneously and were
able to name a shown object correctly, CCT and the
latencies of N20 and P25 had regained baseline levels.
In contrast, the components $ 35 msec remained pro-
longed in latencies and indicated impaired cortical pro-
cessing of sensory stimuli.
In addition, we demonstrated that the MnSSER
latencies P45 and N50 were reduced in the patients
who recalled the immediate wakeup phase after 24 hr
in comparison with the patients without recall. The
total time of anesthesia was longer in those patients
without recall. Presumably the level of anesthesia dif-
fered among the patients due to the persistent uptake
of isoflurane in the different body compartments.
Subsequently, patients had a higher residual of volatile
anesthetics. This was not reflected in the endexpirato-
ry gas concentrations, which did not differ in both
groups. Moreover, the time of extubation and the
time when the patients were able to name a shown
object were similar. It remains open whether a more
detailed clinical investigation including cognitive or
motor tasks performances would have revealed other
differences between the two groups.
(5.6 0.6 msec vs 7.4 1.3 msec,
amplitudes N20P25 and P25N35 decreased significant-
ly ( < 0.001), while the later amplitudes were sup-
P < 0.001). The
P
pressed in almost half of the patients.
During emergence from anesthesia, the MnSSER
changes were gradually reversed. In comparison to
AWAKE, all latencies and CCT remained prolonged (
# 0.03) at Pre-EXT. In comparison to STABLE CCT
(6.6 1.1 msec) and all latencies except N50 ( = 0.06)
were significantly reduced ( # 0.04) at Pre-EXT. N35
P
P
P
was still absent in one and P45 in three patients, while
nine patients did not present N50. The amplitudes
N20P25 and P25N35 remained reduced in comparison
to AWAKE (
with STABLE (
P
< 0.001), while they increased compared
< 0.001). After extubation, when the
P
patients opened their eyes on command (Post-EXT), the
latencies were still prolonged in comparison to AWAKE
(P # 0.05). CCT (5.9 1.1 msec) and the amplitude
P25N35 regained baseline level. The prolongations of all
latencies and CCT were further reduced in comparison
to STABLE (P < 0.01). The N35 remained suppressed
in one patient, P45 in two, and N50 in six patients. At
RECOVERY the latencies of components $ 35 msec
were still prolonged (
AWAKE, while the amplitudes N20P25 (
P45N50 ( = 0.002) remained reduced. P45 was absent
P
# 0.001) in comparison to
P
< 0.02) and
P
in one and N50 in three patients.
MnSSER and memory
At RECOVERY the patients in the group No-MEM
showed prolonged latencies for the components P45
and N50 in comparison to the patients of the group
MEM (P # 0.03; Table III, Figure 3). The amplitudes
and the latencies of N20, P25 and N35 did not differ
between the two groups. During the awake state, sta-
tistical analysis did not reveal differences in the
MnSSER components between the two groups.
Moreover, N20, P25 and N35 did not differ during
anesthesia. The components P45 and N50 were exclud-
ed from the group comparison during STABLE,
because they were suppressed in almost half of the
patients during anesthesia and the number of observa-
tions was too small for statistical analysis. Figure 4 gives
an example of the typical MnSSER traces with the iden-
tification of the various components of one patient who
remembered the MnSSER recording at RECOVERY.
The anesthetic induced MnSSER changes were almost
completely reversed at RECOVERY in this patient.
It is generally accepted that isoflurane prolongs the
latencies and reduces the amplitudes of cortical com-
ponents in a dose-related manner.10,17 Nitrous oxide
has been shown in volunteers to reduce the N20
amplitude without a change in latency and is known to
have an additional effect on isoflurane in depressing
cortical MnSSER components.18,19 In eight patients
Thornton et al. identified a subsequent negative
(N35) and positive deflection (P45) following N20-
P25 complex during anesthesia.12 They documented
statistically different latency changes of N20 and P25
induced by isoflurane compared to nitrous oxide,
while the latencies of two subsequent components
were affected similarly by the two anesthetic agents.
Isoflurane reduced amplitudes of P25-N35 and N35-
P45 more profoundly, while the effect of nitrous oxide
on the amplitude of the P15-N20 and the N20-P25
response was stronger. The difference was statistically
significant for P15-N20 only. The authors suggested

500
CANADIAN JOURNAL OF ANESTHESIA
hepatic encephalopathy for example.2 8While in the pre-
sent study the highpass was chosen according to the
IFCN standard for recording the primary cortical com-
plex N20-P25, for the later components a high pass set-
ting between 1 and 5 Hz is suggested for further
studies.14 Moreover, a lower stimulation frequency
about 1 Hz would be appropriate.
that the effect of nitrous oxide on the subcortical
waves reflected analgesia, while the later cortical waves
were sensitive to the hypnotic potency of isoflurane.
The comparison with our results remains limited due
to differences in methods and statistical analysis. To
support a differential effect of nitrous oxide and
isoflurane a further study with graded changes of the
concentration of one agent at a time is needed.
When interpreting our results, several factors must
be taken into consideration which contribute to minor
changes in evoked responses. Changes in heart rate and
arterial pressure indicated different steps of recovery
from anesthesia clinically, while capnometry reflected
the anesthetic effect on respiratory function. Despite
efforts to maintain body temperature, a decrease of 1C
occurred. However, it has been shown in previous stud-
ies that minor changes in body temperature, hypercar-
bia and blood pressure do not interfere substantially
with evoked response monitoring.20–22 To avoid further
variation, the intensity of stimulation was kept constant
in the present investigation. It has been shown that
increasing stimulation intensity changes the evoked
responses amplitudes.23
In addition, the use of fentanyl and the surgical trau-
ma might have influenced our results. Previous studies
indicated an increase in the amplitudes N20P25 and
P25N35 during surgical periosteal preparation under
isoflurane/nitrous oxide anesthesia.1 1 Unfortunately
there is no systematic study of the effect of small doses of
fentanyl on evoked responses comparing the stimulated
(intubation, skin incision) and the unstimulated patient.
During stable isoflurane anesthesia (0.4 MAC) with
remifentanil, Crabb et al. documented an increase in the
amplitudes of P25N35 and N35P45, while N20P25
decreased with remifentanil in the unstimulated
patient.2 4 It is known from studies with the bispectral
index (derived from the spontaneous electroencephalo-
gram by a complex signal analysing technique), that the
adjunctive use of opioids confounds the interpretation of
the index as a measure of anesthetic adequacy predicting
movement to skin incision.2 5
We showed, in this study, that the later MnSSER
components ($ 35 msec) are particularly sensitive to
different levels of sedation and consecutive cognitive
impairment. Similarly, it is known from the midlatency
auditory evoked responses that Pa and Nb components,
occurring approximately 30 and 40 msec post-stimulus,
are promising markers for changes in consciousness
during anesthesia.26 However, the generators of the
later MnSSER components are still debated and the
nomenclature differs in relation to different recording
techniques.27 However, they have been shown to be
early markers for cognitive impairment in patients with
We were able to document that the anesthetic
induced changes of the MnSSER components P45
and N50 were less pronounced in the patients who
recalled the immediate recovery period 24 hr later.
Further clarification is needed, whether the MnSSER
changes are directly related to impaired memory func-
tion or it is a coincidental finding related to a common
cause. Using a structured interview to assess recall of
the recovery period is a rather global and perhaps less
precise test of memory function. Cognitive functions
are related to different cortical neuronal structures
which are linked to each other by numerous synaptic
processes at the cerebrum. Assessing cerebral function
related to learning and memory in volunteers or
patients is complex.29,30 Memory research investigat-
ing different states of consciousness due to anesthetics
seems to be even more complicated.31 However, the
midlatency components of auditory evoked responses
have been shown to correlate with preservation of
implicit memory function during anesthesia.32 Van
Hooff et al. demonstrated that event-related poten-
tials recorded intraoperatively during cardiac surgery
correlated with postoperative memory.3 3Further stud-
ies are needed to clarify to what extend midlatency
MnSSER components can be used as a tool to evalu-
ate the likelihood of preserved memory function dur-
ing anesthesia.
In conclusion, the present study demonstrated the
practicability and feasibility of recording MnSSER dur-
ing the early recovery period from isoflurane/nitrous
oxide anesthesia. The midlatency MnSSER components
offered information about the reversibility of the anes-
thetic state during the process of clinical awakening.
Concomitantly, the MnSSER latencies differed among
patients with and without recall. It seems worthwhile to
examine MnSSER changes occurring during the various
phases of recovery at subanesthetic concentrations of
isoflurane in order to determine whether there is a cor-
relation with evidence of awareness and whether
MnSSER do supply additional information to the spon-
taneous electroencephalogram or auditory evoked
responses.
References
1 Lopez JR. Intraoperative Neurophysiological
Monitoring. Int Anaesthesiol Clin 1996; 34: 33–54.

Rundshagen et al.: MnSSER DURING RECOVERY
501
15 American Electroencephalographic Society. Guidelines
for clinical evoked potential studies. J Clin
Electrophysiol 1984; 1: 3–53.
16 Clark DL, Rosner BS. Neurophysiologic effects of gen-
eral anesthestics: I. The electroencephalogram and sen-
sory evoked responses in Man. Anesthesiology 1973;
38: 564–82.
17 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.
18 Sebel PS, Flynn PJ, Ingram DA. Effect of nitrous oxide
on visual, auditory and somatosensory evoked poten-
tials. Br J Anaesth 1984; 56: 1403–7.
19 Lam AM, Sharar SR, Mayberg TS, Eng CC. Isoflurane
compared with nitrous oxide anaesthesia for intraoper-
ative monitoring of somatosensory-evoked potentials.
Can J Anaesth 1994; 41: 295–300.
20 Porkkala T, Kaukinen S, Häkkinen V, Jäntti V. Effects
of hypothermia and sternal retractors on median nerve
somatosensory evoked potentials. Acta Anaesthesiol
Scand 1997; 41: 843–8.
21 Ledsome JR, Cole C, Sharp-Kehl JM. Somatosensory
evoked potentials during hypoxia and hypocapnia in
conscious humans. Can J Anaesth 1996; 43: 1025–9.
22 Kalkman CJ, Boezeman EH, Ribberink AA, Oosting J,
Deen L, Bovill JG. Influence of changes in arterial car-
bon dioxide tension on the electroencephalogram and
posterior tibial nerve somatosensory cortical evoked
potentials during alfentanil/nitrous oxide anesthesia.
Anesthesiology 1991; 75: 68–74.
23 Lesser RP, Koehle R, Lueders H. Effect of stimulus inten-
sity on short latency somatosensory evoked potentials.
Electroenceph Clin Neurophysiol 1979; 47: 377–82.
24 Crabb I, Thornton C, Konieczko KM, et al. Remifentanil
reduces auditory and somatosensory evoked responses
during isoflurane anaesthesia in a dose-dependant man-
ner. Br J Anaesth 1996; 76: 795–801.
25 Sebel PS, Lang E, Rampil IJ, et al. A multicenter study of
bispectral electroencephalogram analysis for monitoring
anesthetic effect. Anesth Analg 1997; 84: 891–9.
26 Davies FW, Mantzaridis H, Kenny GNC, Fisher AC.
Middle latency auditory evoked potentials during
repeated transitions from consciousness to uncon-
sciousness. Anaesthesia 1996; 51: 107–13.
2 Stöhr M. Somatosensible Reizantworten von
Rückenmark und Gehirn. In: Stöhr M, Dichgans J,
Buettner UW, Hess ChW, Altenmüller E (Eds.).
Evozierte Potentiale, 3rd ed. Berlin: Springer Verlag,
1996: 23–129.
3 Allison T, McCarthy G, Wood CC, Jones SJ. Potentials
evoked in human and monkey cerebral cortex by stim-
ulation of the median nerve. A review of scalp and
intracranial recordings. Brain 1991; 114: 2465–503.
4 Yamada T. The anatomic and physiologic bases of the
median nerve somatosensory evoked potentials. Neurol
Clin 1988; 6: 705–33.
5 Allison T, McCarthy G, Wood CC. The relationship
between human long-latency somatosensory evoked
potentials recorded from the cortical surface and from
the scalp. Electroenceph Clin Neurophysiol 1992; 84:
301–14.
6 Srisa-an P, Lei L, Tarkka IM. Middle latency
somatosensory evoked potentials: noninvasive source
analysis. J Clin Neurophysiol 1996; 13: 156–63.
7 Baumgartner C, Sutherling WW, Di S, Barth DS.
Spatiotemporal modeling of cerebral evoked magnetic
fields to median nerve stimulation. Electroenceph Clin
Neurophysiol 1991; 79: 27–35.
8 Baumgartner C, Doppelbauer A, Deecke L, et al.
Neuromagnetic investigation of somatotopy of human
hand somatosensory cortex. Exp Brain Res 1991; 87:
641–8.
9 Sutherling WW, Levesque MF, Baumgartner C. Cortical
sensory representation of the human hand: size of fin-
ger regions and nonoverlapping digit somatotopy.
Neurology 1992; 42: 1020–8.
10 Samra SK, Vanderzant CW, Domer PA, Sackellares JC.
Differential effects of isoflurane on human median
nerve somatosensory evoked potentials. Anesthesiology
1987; 66: 29–35.
11 Rundshagen I, Kochs E, Schulte am Esch J. Surgical
stimulation increases median nerve somatosensory
evoked responses during isoflurane-nitrous oxide anaes-
thesia. Br J Anaesth 1995; 75: 598–602.
12 Thornton C, Creach-Barry P, Jordan C, et al.
Somatosensory and auditory evoked responses record-
ed simultaneously: differential effects of nitrous oxide
and isoflurane. Br J Anaesth 1992; 68: 508–14.
13 De Beer NAM, van Hooff JC, Brunia CHM, Cluitmans
PJM, Korsten HHM, Beneken JEW. Midlatency auditory
evoked potentials as indicators of perceptual processing
during general anaesthesia. Br J Anaesth 1996; 77:
617–24.
27 Desmedt JE. Somatosensory evoked potentials. In:
Picton TW (Ed.). Handbook of
Electroencephalography and Clinical Neurophysiology,
Vol. 3. Amsterdam: Elsevier, 1988: 245–340.
28 Madl C, Grimm G, Ferenci P, et al. Serial recording of
sensory evoked pontentials: a noninvasive prognostic
indicator in fulminant liver failure. Hepatology 1994;
20: 1487–94.
14 Nuwer MR, Aminoff M, Desmedt J, et al. IFCN recom-
mended standards for short latency somatosensory
evoked potentials. Report of an IFCN committee.
Electroenceph Clin Neurophyisol 1994; 91: 6–11.

502
CANADIAN JOURNAL OF ANESTHESIA
29 Andrade J, Baddeley A. Human memory and anesthe-
sia. Int Anesthesiol Clin 1993; 31: 39–51.
30 Ghoneim MM, Block RI. Learning during anesthesia.
Int Anesthesiol Clin 1993; 31: 53–65.
31 Jelicic M, Bonke B. Learning during anaesthesia
(Editorial). Eur J Anaesthesiol 1996; 13: 311–3.
32 Schwender D, Kaiser A, Klasing S, Peter K, Pöppel E.
Midlatency auditory evoked potentials and explicit and
implicit memory in patients undergoing cardiac
surgery. Anesthesiology 1994; 80: 493–501.
33 Van Hooff JC, De Beer NAM, Brunia CHM, et al.
Information processing during cardiac surgery: an
event related potential study. Electroenceph Clin
Neurophysiol 1995; 96: 433–52.