Resuscitation, 23 (1992) 235-248
235
Elsevier Scientific Publishers Ireland Ltd.
Evaluation of the central nervous function in resuscitated comatose patients by multilevel evoked potentials Tatsuhiko Kanoa, Osamu Shimodab, Tohru Moriokac, Yoshihiro Yagishitad and Akira Hashiguchid “Surgical Center, Kumamoto University Hospital, Kumamoto 860, bDepartment of Anesthesiology, Kumamoto Rosai Hospital, Yatsushiro 866, ‘Department of Anesthesiology, Kumamoto University Hospital, Kumamoto 860 and dDepartment of Anesthesiology, National Hospital Medical Center, Tokyo 162 (Japan)
(Received April lst, 1992; accepted May 14th, 1992)
Multilevel evoked potentials were examined in 17 patients who became comatose after cardiac arrest and resuscitation. In 4 patients, the P, through N, components of the somatosensory evoked cerebral potential (SECP) were present altogether within 100 ms after the ischemic insults. They all subsequently regained consciousness, though three of them developed intelligence and motor disturbances to some extent. In 11 patients who regained consciousness, or remained in a vegetative state, the evoked potentials which reflect brainstem functions, such as the auditory evoked brainstem potential, the R, wave of the orbicularis oculi reflex and the slow positive wave of the somatosensory evoked brainstem potential, were recognized. The somatosensory evoked spinal potential and spinal monosynaptic reflex showed normal appearances in the state of vegetation and even after the determination of brain death. The measures of SECP could be useful in predicting restoration of consciousness. Key words: somatosensory evoked potential; auditory evoked potential; cardiopulmonary resuscitation; postischemic brain damage
INTRODUCTION
The development of computer technology has resulted in the simplification and extensive clinical use of brain function tests using evoked potentials. Evoked potential tests are being actively used as a diagnostic modality for brain death [1,2], the intraoperative monitoring of brain function [3,4] and as an indicator in evaluating the severity and prognosis of brain damage in comatose patients with head trauma [5-71 or ischemic insult [6,7], or toxic-metabolic encephalopathy [7,8]. We have been performing evoked potential tests at the cerebral, brainstem and spinal cord levels to evaluate the function of the central nervous system (CNS) in comatose patients who have been resuscitated from cardiac arrest [9,10]. Based on these findings, this report describes the usefulness and limitations of multi level evoked potentials in determining the severity and prognosis of CNS damage and cautions in conducting these tests. Correspondence to: Tatsuhiko Kano, Vice Director, Surgical Center, Kumamoto University Hospital, I-I-
I Honjo, Kumamoto 860, Japan. 0300-9572/92/%05.00 0 1992 Elsevier Scientific Publishers Ireland Ltd.
236 SUBJECTS AND METHODS
Of the comatose patients who were resuscitated from cardiac arrest at our hospital or transported to our hospital after surviving cardiopulmonary resuscitation between 1974 and 1991, 17 who had undergone evoked potential tests, were evaluated (Table I). This investigation was approved from the Institutional Ethics Committee on Human Research at the start in 1974. Informed consent was obtained from spouse, parents or close relatives of the patients. Evoked potentials were examined at the bedside by several methods. (I) Somatosensory evoked cerebral potential (SECP) The median nerve near the wrist joint was stimulated with a rectangular electric current (0.3 ms in pulse width, 3-5 mA in intensity and l-2 Hz in frequency) and the SECP was recorded from the contralateral post-Rolandic area with a reference electrode at the ear lobe. Needle electrodes were used for stimulation and recording. Band-pass filters were set at lOO- 1500 Hz. Additive averaging was repeated 100-200 times and upward deflections were displayed as negative waves. Indication of the component waves by latencies was avoided, because the latency of each component wave of SECP in these patients was often far from the physiological criteria. Instead, we followed the conventional nomenclature: the first small positive wave was termed P, and the subsequent alternate positive and negative waves Nt, P2, Nz, PJ.... The latency of each component was presented by the peak latency. To avoid interference by myogenic components in the SECP test, the patients were immobilized by the administration of muscle relaxants, with respiration controlled. (2) Somatosensory evoked brainstem potential (SEBP) The median nerve was stimulated similarly to the SECP test and the SEBP was recorded unipolarly or bipolarly through a stainless steel needle electrode (0.3 mm in diameter and 5 cm in length and insulated except for the tip) which was vertically inserted into the pharyngeal nodule of the basilar part of the occipital bone through the median part of the soft palate [ 11,121. Prior to the placement of the needle electrode, the oropharyngeal cavity was disinfected with 0.025% (w/v) benzethonium chloride. The patients undergoing SEBP tests were also given a long acting muscle relaxant, pancuronium, to ensure safety and to avoid interference by myogenic components. Setting of the band-pass filter, the frequency of additive averaging and the style of wave form indication were the same as those for SECP. (3) Auditory evoked brainstem potential (AEBP) A click sound (250 p in wavelength, 100 dB in sound pressure and lo-20 Hz in frequency) was given to both ears through a headphone for vertex-mastoid process recording. Before starting the AEBP test, esophageal, rectal or blood temperature was measured and environmental periodic sounds, produced by an ECG monitor, a ventilator and so on, were reduced as much as to be a minimum. The band-pass filter was set at 100-1500 Hz. Additive averaging was performed 1000 or 2000 times and upward deflections were indicated as positive waves.
Asthma attack Drowning
M M
I7
1
Asthma attack Drowning Drowning Myocardial infarction Circulatory shock Myocardial infarction Strangulation
F M F F F M M M
36
59 2 61
63
78
51
22
IO
II I2 13
14s
lSC
16c
l7d
+
+*
+
+ +* +’
+
+*
+
+
“Survived more than 3 months in a state of semi-coma or coma. bSevere intellectual and motor disturbances were observed. cSlight intellectual and motor disturbances were observed. dNo neurological disturbance observed. *Under pentobarbital and mild hypothermia therapy. **Disappeared on day 2. ***Disappeared on day 5.
Circulatory shock
Strangulation
F
22
9
1.7
+
Asphyxia
M
41
6
8
+
+ +*
Idiopatic pneumothorax Drug induced shock Asthma attack Hypoxic hypoxia
M M F F
??
??
??
??
??
???
??? ? ?
??
??
??
??
??
??
??
??
??
??
??
??
??
??
??
??
??
??
??? ?
??? ? ?
??
??
??
??
??
??
2 8 I 2 3 5 I
I 2 I 5 I 2 I 2 I8 2 I
2 3 2 2 2 I 5 2
-
_* ??? ?
Day
LR
Convulsion
CR
EEG
Early neurological features
43 50 68 43
2 3 4 5
Cause of cardiac arrest
Cardiac tamponade
32
1
Sex
F
Age (yr)
Patient
a Wave 6 Wave e Wave 6 Wave (Y - j3 Wave 6 - OLWave
e-
6 Wave
6 Wave Spike and 6 wave 6 Wave Spike and 6 wave Almost isoelectric 6 Wave 6 Wave Burst and suppression Spike and 6 wave Spike and d wave 6 Wave
Spike and suppression Isoelectric Isoelectric Spike and 6 wave Almost isoelectric Spike and suppression Isoelectric Spike and suppression
Feature
on on on on
day day day day
5 2 IO I8
Restored consciousness on day I6 Restored consciousness on day I Restored consciousness on day 5 Restored consciousness on day 2
Vegetative statea Vegetative states Vegetative state”
Vegetative state”
Vegetative state”
Vegetative state”
Vegetative states
Brain death on day IO
Died Died Died Died
Died on day 7
Outcome
Table I. Seventeen comatose patients undergoing examination of evoked potentials after resuscitation from cardiac arrests. LR, light reaction; CR Cornea1 reaction; Day, and on day, denote number of days after resuscitation.
238
(4) Orbicularis oculi rejlex (OOR)
The unilateral forehead innervated by the supraorbital nerve was stimulated with a rectangular electric current (0.1 ms in pulse width and 5-7 mA in intensity) and the reflex electromyogram (EMG) was recorded bilaterally from the skin above the orbicular oculi muscle below the orbit and external canthus [13]. Surface disk electrodes were used for stimulation and recording, To prevent a conditioning experience, stimulation was applied randomly at intervals of 5 s or more. OOR was superimposed three times and recorded for analysis. (5) Somatosensory evoked spinal potential (SESP)
and spinal monosynaptic rejlex
(SMR)
The posterior tibia1 nerve at the popliteal fossa was stimulated with a rectangular electric current (0.3 ms in pulse width and l-2 Hz in frequency) for recording of segmental SESP from the Th1i-Th12 or Th12-L, epidural space and nerve action potential of the afferent and efferent spinal roots forming SMR from the lumbar subdural space. A polyethylene tube (0.9 mm in diameter) was first introduced into the epidural space or subdural space through an 18-gauge Tuohy needle. Then, a stainless-steel recording wire electrode, 125 pm in diameter, insulated except for 5 mm of the tip, was advanced through the tube until the tip of the wire electrode appeared from the tube [14]. The reference needle electrode was inserted into the adjacent supraspinal ligament. Additive averaging was repeated 20-50 times and upward deflections were indicated as positive waves. The evoked electromyogram (EEMG) was simultaneously recorded from the calf muscle on the stimulated side. Patients were judged to be in a vegetative state if they continued to show impaired consciousness for more than 3 months after resuscitation from cardiac arrest, accompanied by inabilities of communicating more than a weak verbal response to a simple order, movement and eating and urinary and fecal incontinence. Until 1984, brain death was judged according to the criteria proposed by the Japan Association of Electroencephalography in 1974 [15], though an indication of the criteria was limited to primary gross lesions. Since 1985, the determination of brain death was based on the guidelines and criteria established by the Japan Ministry of Health and Welfare in 1985 [16], which covered secondary CNS lesions. RESULTS
(1) Cerebrum SECP. The early components up to Ns were not present together within 100 ms
of latency in patients 1 and 4 who fell into brain death without regaining consciousness, and patients 7, 9- 11 and 13 who subsequently remained in a vegetative state (Fig. la). Patients 1 and 4 exhibited none of the components subsequent to the initial dip. In four patients who regained consciousness and the ability to speak (patients 14-17), the latencies of the late components were extended, but the early components, Pi-Ns, were all present within 100 ms of latency in early tests after resuscitation (Fig. lb). Patient 14, who recovered consciousness on day 16 after resuscitation, showed a temporary reduction of the amplitudes of early SECP components on proximity to recovery of consciousness (Fig. 2) [9]. On serial examina-
239 (a) 02
Patient l10
o4
4
-
15
p2
D-6
16
24
16 22
36 20
a o-
‘7
msec
1
58 20
n*ec
Fig. 1. Presence and absence of recovery of consciousness and SECP: (a) schemes of SECP in seven patients who did not recover consciousness. Patients 1and 4 went into brain death several days later and died, patients 7, 9, 10, 11and 13 became vegetative (D-, denotes the number of days after resuscitation from cardiac arrest, numbers on the wave forms represent the peak latencies of the respective components (ms)); (b) schemes of SECP in four patients who recovered consciousness, response to verbal order appeared on day 16 in patient 14, day 7 in patient 15, day 5 in patient 16 and day 2 in patient 17. The standard latencies in Japanese adults [17] (in ms) are: P,, 13.50 f 1.20 (S.D.); N,, 18.27 f 1.06; P2, 22.98 f 1.99; N,, 30.4 i 2.65; P,, 39.61 f 3.23; Ns, 53.25 * 3.96.
240
AEEP
SECP
12.0 Y”
Fig. 2. Serial changes of SECP and AEBP in patient 14 in the course of recovering consciousness. The patient came to respond to simple verbal orders soon after the examination on day 16. Shortly before she regained consciousness, each component of the AEBP had shown prolonging in latency and decreasing in amplitude and the components of the SECP also had decreased in amplitude. At the examination done 2 weeks after the restoration of consciousness, the AEBP and the SECP configurations showed a tendency to recovery. Her body temperature was within the range of 36.3-37.8”C during the period and the stimulus intensity was not altered.
tions in a state of vegetation, the SECP configurations showed only slight improvement (Fig. 3). In the patients showing marked myotonia, myogenic components entered the SECP before administration of a muscle relaxant, interfering with differentiation between the two types of components (Fig. 4). (2) Brainstem SEBP. The SEBP of patients, who regained consciousness or were in a vegetative state, consisted of sharp positive and negative waves and then, a slow positive wave of 20-30 ms. In contrast, the SEBP of patients who deteriorated into brain death next day was devoid of slow positive waves (Fig. 5). AEBP. In some of the patients, who regained consciousness or were in a vegetative state, the AEBP showed a slight delay of latency, but clearly demonstrated the Vth wave (Table II). The latency showed slight changes during the follow-up period and within the same day. The relationship between the anticonvulsant, administered for suppression of convulsion, such as diazepam, pentobarbital and phenytoin and the extension of the latency of AEBP was unknown. Patient 14 exhibited temporary
241
D-6
e-j---J--
71
D-14
J\
-1
40 msec
20 Fig. 3. Changes in SECP in patient IO who resuscitation from cardiac arrest. After a lapse There was disorientation and dyscalculia. She cian. The periphery of the extremities showed
became vegetative D-, denotes the number of days after of 3 years, ‘yes’ was the only word the patient could speak. was barely able to distinguish her mother from the physicontracture and marked myotonia.
EEG
Lt. F-O
Ht.
F-O
30
msec
Fig. 4. Contamination of myogenic components to SECP in patient 7. SECP test was performed on day I3 after resuscitation. The SECP configurations recorded simultaneously from the contra- and ipsilateral post-Rolandic areas disappeared except the early negative dip after the administration of succinylcholine chloride (SCC). The upper traces show EEGs after the SCC administration.
242
SECP
Fig. 5. SEBP in vegetative state and deteriorating into brain death. SEBP was recorded with the epipharynx (Eph)-subocciput (SO) bipolar lead in response to electric stimulation of the median nerve under pancuronium administration and SECP was recorded with the scalp (S)-ear (E) monopolar lead, simultaneously. Left column: SECP, SEBP and EEC of patient 10 who remained in a vegetative state. Right column: SECP and SEBP of patient 1 (day 2) who fell into brain death. Lower trace shows EEG at an evoked potential test. EEG became flat on day 3 and the patient died on day 7.
extension and a decrease in amplitude of the latency of the Vth wave immediately before recovery of consciousness (Fig. 2) [9]. Patient 6, who deteriorated into brain death, exhibited a normal appearance of AEBP on the day of resuscitation, delay of latencies on the second day and none of the components of AEBP on the third day. OOR. Even in the patients who regained consciousness, the OOR contained only the R, wave (Table III). In patient 14, who regained consciousness, the R2 wave, with a delayed latency appeared in the test on day 14 after resuscitation and subsequently, the latency was gradually shortened [9]. Both the RI and R2 waves were present even after transition to a vegetative state in some patients, but the latter wave was sharply triphasic (Fig. 6), rather than polyphasic, a common form. Neither the RI nor R2 waves were present in the patients who received diazepam or pentobarbital and in those who became brain dead several days after resuscitation. (3) Spinal Cord SESP, SMR and EEMG (M and H waves). The SESP showed a completely normal pattern even in a vegetative state with markedly disturbed SECP. Examination of conduction changes in the reflex arc of SMR in response to a gradual increase in the stimulating intensity in terms of action potentials of the cauda equina in the subarachnoid space of the lumbar area and H waves of the calf muscle revealed a pattern similar to that in normal individuals [19] (Fig. 7).
243 Table II.
Latency of Vth wave of AEBP. Patients 4 and 6 deteriorated into brain death, patients 9-13 remained in a vegetative state, and patients 14, 15 and 17 recovered consciousness. The standard latency of the Vth wave of AEBP in Japanese adults is 5.65 * 1.0 ms (see Ref. 18). Patient
4 6
9 10
11
12 13 14
No. of days after resuscitation
Latency of Vth wave (ms)
4 la 2” 3 2oa 27a 9” 11” 13” 16a 50” 62” 111” 28
5.3 6.7 5.6 5.6-5.7 5.4 5.5 5.3-5.4 5.5 6.5 6.3 6.2 6.3 5.9 6.3 6.3 6.3 8.4 8.5 6.3 5.4 5.4 6.1
2” 5 9 14 16 29 la 3 1
15 17
‘Under administration of anticonvulsants (diazepam, pentobarbital and/or phenytoin). Table IN.
Appearance of Rt and R, waves of the OOR and simultaneously recorded EEGs. Patients 2 and 5 deteriorated into brain death, patients 8-10 remained in a vegetative state, and patients 14 and 16 recovered consciousness.
Patient
No. of days after resuscitation
OOR RI
2 5 8 9 10 14
16
2 5 81 lga 3a 9 14 16 29 6
EEG R2
+*
+ + +
Isoelectric Isoelectric Spike and fl wave Spike and 6 wave 0 Wave Low voltage 6 wave Low voltage 6 wave Low voltage 0 wave Low voltage 0 wave ci wave
‘Under treatment with anticonvulsants (diazepam, pentobarbital and/or phenytoin). *R2 wave was sharply triphasic.
244 OOR
Rl
fi
h
R2
EEG
1
see
Fig. 6. Sharp triphasic R, wave in vegetated patient 8 (day 81) with increased muscle tone. The lower traces show EEGs recorded monopolarly at that time.
DISCUSSION
In general, disturbance of the CNS following systemic ischemia or anoxia, such as resuscitation from cardiac arrest, extends gradually from the cerebrum, which has weak resistance, to the spinal cord, depending on the severity [20]. For comatose patients immediately after resuscitation from cardiac arrest, objective evaluation of the cerebral function is required. Although the severity of disturbance of the CNS and the patient’s prognosis can be roughly estimated if the interval between cardiac arrest and resuscitation is known, accurate determination of the duration of circulatory arrest is frequently difficult. Furthermore, since many other factors including the situation in which cardiac arrest occurred, underlying diseases and age are also involved, it is difficult to judge the severity and prognosis on the basis of the time factor alone. In this study of evoked potentials, SECP and R2 wave of OOR were used as indicators of cerebral function. The patients who showed all of the early components
245 3
mA
(La-L,)",7;EEMG
235
Elsevier Scientific Publishers Ireland Ltd.
Evaluation of the central nervous function in resuscitated comatose patients by multilevel evoked potentials Tatsuhiko Kanoa, Osamu Shimodab, Tohru Moriokac, Yoshihiro Yagishitad and Akira Hashiguchid “Surgical Center, Kumamoto University Hospital, Kumamoto 860, bDepartment of Anesthesiology, Kumamoto Rosai Hospital, Yatsushiro 866, ‘Department of Anesthesiology, Kumamoto University Hospital, Kumamoto 860 and dDepartment of Anesthesiology, National Hospital Medical Center, Tokyo 162 (Japan)
(Received April lst, 1992; accepted May 14th, 1992)
Multilevel evoked potentials were examined in 17 patients who became comatose after cardiac arrest and resuscitation. In 4 patients, the P, through N, components of the somatosensory evoked cerebral potential (SECP) were present altogether within 100 ms after the ischemic insults. They all subsequently regained consciousness, though three of them developed intelligence and motor disturbances to some extent. In 11 patients who regained consciousness, or remained in a vegetative state, the evoked potentials which reflect brainstem functions, such as the auditory evoked brainstem potential, the R, wave of the orbicularis oculi reflex and the slow positive wave of the somatosensory evoked brainstem potential, were recognized. The somatosensory evoked spinal potential and spinal monosynaptic reflex showed normal appearances in the state of vegetation and even after the determination of brain death. The measures of SECP could be useful in predicting restoration of consciousness. Key words: somatosensory evoked potential; auditory evoked potential; cardiopulmonary resuscitation; postischemic brain damage
INTRODUCTION
The development of computer technology has resulted in the simplification and extensive clinical use of brain function tests using evoked potentials. Evoked potential tests are being actively used as a diagnostic modality for brain death [1,2], the intraoperative monitoring of brain function [3,4] and as an indicator in evaluating the severity and prognosis of brain damage in comatose patients with head trauma [5-71 or ischemic insult [6,7], or toxic-metabolic encephalopathy [7,8]. We have been performing evoked potential tests at the cerebral, brainstem and spinal cord levels to evaluate the function of the central nervous system (CNS) in comatose patients who have been resuscitated from cardiac arrest [9,10]. Based on these findings, this report describes the usefulness and limitations of multi level evoked potentials in determining the severity and prognosis of CNS damage and cautions in conducting these tests. Correspondence to: Tatsuhiko Kano, Vice Director, Surgical Center, Kumamoto University Hospital, I-I-
I Honjo, Kumamoto 860, Japan. 0300-9572/92/%05.00 0 1992 Elsevier Scientific Publishers Ireland Ltd.
236 SUBJECTS AND METHODS
Of the comatose patients who were resuscitated from cardiac arrest at our hospital or transported to our hospital after surviving cardiopulmonary resuscitation between 1974 and 1991, 17 who had undergone evoked potential tests, were evaluated (Table I). This investigation was approved from the Institutional Ethics Committee on Human Research at the start in 1974. Informed consent was obtained from spouse, parents or close relatives of the patients. Evoked potentials were examined at the bedside by several methods. (I) Somatosensory evoked cerebral potential (SECP) The median nerve near the wrist joint was stimulated with a rectangular electric current (0.3 ms in pulse width, 3-5 mA in intensity and l-2 Hz in frequency) and the SECP was recorded from the contralateral post-Rolandic area with a reference electrode at the ear lobe. Needle electrodes were used for stimulation and recording. Band-pass filters were set at lOO- 1500 Hz. Additive averaging was repeated 100-200 times and upward deflections were displayed as negative waves. Indication of the component waves by latencies was avoided, because the latency of each component wave of SECP in these patients was often far from the physiological criteria. Instead, we followed the conventional nomenclature: the first small positive wave was termed P, and the subsequent alternate positive and negative waves Nt, P2, Nz, PJ.... The latency of each component was presented by the peak latency. To avoid interference by myogenic components in the SECP test, the patients were immobilized by the administration of muscle relaxants, with respiration controlled. (2) Somatosensory evoked brainstem potential (SEBP) The median nerve was stimulated similarly to the SECP test and the SEBP was recorded unipolarly or bipolarly through a stainless steel needle electrode (0.3 mm in diameter and 5 cm in length and insulated except for the tip) which was vertically inserted into the pharyngeal nodule of the basilar part of the occipital bone through the median part of the soft palate [ 11,121. Prior to the placement of the needle electrode, the oropharyngeal cavity was disinfected with 0.025% (w/v) benzethonium chloride. The patients undergoing SEBP tests were also given a long acting muscle relaxant, pancuronium, to ensure safety and to avoid interference by myogenic components. Setting of the band-pass filter, the frequency of additive averaging and the style of wave form indication were the same as those for SECP. (3) Auditory evoked brainstem potential (AEBP) A click sound (250 p in wavelength, 100 dB in sound pressure and lo-20 Hz in frequency) was given to both ears through a headphone for vertex-mastoid process recording. Before starting the AEBP test, esophageal, rectal or blood temperature was measured and environmental periodic sounds, produced by an ECG monitor, a ventilator and so on, were reduced as much as to be a minimum. The band-pass filter was set at 100-1500 Hz. Additive averaging was performed 1000 or 2000 times and upward deflections were indicated as positive waves.
Asthma attack Drowning
M M
I7
1
Asthma attack Drowning Drowning Myocardial infarction Circulatory shock Myocardial infarction Strangulation
F M F F F M M M
36
59 2 61
63
78
51
22
IO
II I2 13
14s
lSC
16c
l7d
+
+*
+
+ +* +’
+
+*
+
+
“Survived more than 3 months in a state of semi-coma or coma. bSevere intellectual and motor disturbances were observed. cSlight intellectual and motor disturbances were observed. dNo neurological disturbance observed. *Under pentobarbital and mild hypothermia therapy. **Disappeared on day 2. ***Disappeared on day 5.
Circulatory shock
Strangulation
F
22
9
1.7
+
Asphyxia
M
41
6
8
+
+ +*
Idiopatic pneumothorax Drug induced shock Asthma attack Hypoxic hypoxia
M M F F
??
??
??
??
??
???
??? ? ?
??
??
??
??
??
??
??
??
??
??
??
??
??
??
??
??
??
??
??? ?
??? ? ?
??
??
??
??
??
??
2 8 I 2 3 5 I
I 2 I 5 I 2 I 2 I8 2 I
2 3 2 2 2 I 5 2
-
_* ??? ?
Day
LR
Convulsion
CR
EEG
Early neurological features
43 50 68 43
2 3 4 5
Cause of cardiac arrest
Cardiac tamponade
32
1
Sex
F
Age (yr)
Patient
a Wave 6 Wave e Wave 6 Wave (Y - j3 Wave 6 - OLWave
e-
6 Wave
6 Wave Spike and 6 wave 6 Wave Spike and 6 wave Almost isoelectric 6 Wave 6 Wave Burst and suppression Spike and 6 wave Spike and d wave 6 Wave
Spike and suppression Isoelectric Isoelectric Spike and 6 wave Almost isoelectric Spike and suppression Isoelectric Spike and suppression
Feature
on on on on
day day day day
5 2 IO I8
Restored consciousness on day I6 Restored consciousness on day I Restored consciousness on day 5 Restored consciousness on day 2
Vegetative statea Vegetative states Vegetative state”
Vegetative state”
Vegetative state”
Vegetative state”
Vegetative states
Brain death on day IO
Died Died Died Died
Died on day 7
Outcome
Table I. Seventeen comatose patients undergoing examination of evoked potentials after resuscitation from cardiac arrests. LR, light reaction; CR Cornea1 reaction; Day, and on day, denote number of days after resuscitation.
238
(4) Orbicularis oculi rejlex (OOR)
The unilateral forehead innervated by the supraorbital nerve was stimulated with a rectangular electric current (0.1 ms in pulse width and 5-7 mA in intensity) and the reflex electromyogram (EMG) was recorded bilaterally from the skin above the orbicular oculi muscle below the orbit and external canthus [13]. Surface disk electrodes were used for stimulation and recording, To prevent a conditioning experience, stimulation was applied randomly at intervals of 5 s or more. OOR was superimposed three times and recorded for analysis. (5) Somatosensory evoked spinal potential (SESP)
and spinal monosynaptic rejlex
(SMR)
The posterior tibia1 nerve at the popliteal fossa was stimulated with a rectangular electric current (0.3 ms in pulse width and l-2 Hz in frequency) for recording of segmental SESP from the Th1i-Th12 or Th12-L, epidural space and nerve action potential of the afferent and efferent spinal roots forming SMR from the lumbar subdural space. A polyethylene tube (0.9 mm in diameter) was first introduced into the epidural space or subdural space through an 18-gauge Tuohy needle. Then, a stainless-steel recording wire electrode, 125 pm in diameter, insulated except for 5 mm of the tip, was advanced through the tube until the tip of the wire electrode appeared from the tube [14]. The reference needle electrode was inserted into the adjacent supraspinal ligament. Additive averaging was repeated 20-50 times and upward deflections were indicated as positive waves. The evoked electromyogram (EEMG) was simultaneously recorded from the calf muscle on the stimulated side. Patients were judged to be in a vegetative state if they continued to show impaired consciousness for more than 3 months after resuscitation from cardiac arrest, accompanied by inabilities of communicating more than a weak verbal response to a simple order, movement and eating and urinary and fecal incontinence. Until 1984, brain death was judged according to the criteria proposed by the Japan Association of Electroencephalography in 1974 [15], though an indication of the criteria was limited to primary gross lesions. Since 1985, the determination of brain death was based on the guidelines and criteria established by the Japan Ministry of Health and Welfare in 1985 [16], which covered secondary CNS lesions. RESULTS
(1) Cerebrum SECP. The early components up to Ns were not present together within 100 ms
of latency in patients 1 and 4 who fell into brain death without regaining consciousness, and patients 7, 9- 11 and 13 who subsequently remained in a vegetative state (Fig. la). Patients 1 and 4 exhibited none of the components subsequent to the initial dip. In four patients who regained consciousness and the ability to speak (patients 14-17), the latencies of the late components were extended, but the early components, Pi-Ns, were all present within 100 ms of latency in early tests after resuscitation (Fig. lb). Patient 14, who recovered consciousness on day 16 after resuscitation, showed a temporary reduction of the amplitudes of early SECP components on proximity to recovery of consciousness (Fig. 2) [9]. On serial examina-
239 (a) 02
Patient l10
o4
4
-
15
p2
D-6
16
24
16 22
36 20
a o-
‘7
msec
1
58 20
n*ec
Fig. 1. Presence and absence of recovery of consciousness and SECP: (a) schemes of SECP in seven patients who did not recover consciousness. Patients 1and 4 went into brain death several days later and died, patients 7, 9, 10, 11and 13 became vegetative (D-, denotes the number of days after resuscitation from cardiac arrest, numbers on the wave forms represent the peak latencies of the respective components (ms)); (b) schemes of SECP in four patients who recovered consciousness, response to verbal order appeared on day 16 in patient 14, day 7 in patient 15, day 5 in patient 16 and day 2 in patient 17. The standard latencies in Japanese adults [17] (in ms) are: P,, 13.50 f 1.20 (S.D.); N,, 18.27 f 1.06; P2, 22.98 f 1.99; N,, 30.4 i 2.65; P,, 39.61 f 3.23; Ns, 53.25 * 3.96.
240
AEEP
SECP
12.0 Y”
Fig. 2. Serial changes of SECP and AEBP in patient 14 in the course of recovering consciousness. The patient came to respond to simple verbal orders soon after the examination on day 16. Shortly before she regained consciousness, each component of the AEBP had shown prolonging in latency and decreasing in amplitude and the components of the SECP also had decreased in amplitude. At the examination done 2 weeks after the restoration of consciousness, the AEBP and the SECP configurations showed a tendency to recovery. Her body temperature was within the range of 36.3-37.8”C during the period and the stimulus intensity was not altered.
tions in a state of vegetation, the SECP configurations showed only slight improvement (Fig. 3). In the patients showing marked myotonia, myogenic components entered the SECP before administration of a muscle relaxant, interfering with differentiation between the two types of components (Fig. 4). (2) Brainstem SEBP. The SEBP of patients, who regained consciousness or were in a vegetative state, consisted of sharp positive and negative waves and then, a slow positive wave of 20-30 ms. In contrast, the SEBP of patients who deteriorated into brain death next day was devoid of slow positive waves (Fig. 5). AEBP. In some of the patients, who regained consciousness or were in a vegetative state, the AEBP showed a slight delay of latency, but clearly demonstrated the Vth wave (Table II). The latency showed slight changes during the follow-up period and within the same day. The relationship between the anticonvulsant, administered for suppression of convulsion, such as diazepam, pentobarbital and phenytoin and the extension of the latency of AEBP was unknown. Patient 14 exhibited temporary
241
D-6
e-j---J--
71
D-14
J\
-1
40 msec
20 Fig. 3. Changes in SECP in patient IO who resuscitation from cardiac arrest. After a lapse There was disorientation and dyscalculia. She cian. The periphery of the extremities showed
became vegetative D-, denotes the number of days after of 3 years, ‘yes’ was the only word the patient could speak. was barely able to distinguish her mother from the physicontracture and marked myotonia.
EEG
Lt. F-O
Ht.
F-O
30
msec
Fig. 4. Contamination of myogenic components to SECP in patient 7. SECP test was performed on day I3 after resuscitation. The SECP configurations recorded simultaneously from the contra- and ipsilateral post-Rolandic areas disappeared except the early negative dip after the administration of succinylcholine chloride (SCC). The upper traces show EEGs after the SCC administration.
242
SECP
Fig. 5. SEBP in vegetative state and deteriorating into brain death. SEBP was recorded with the epipharynx (Eph)-subocciput (SO) bipolar lead in response to electric stimulation of the median nerve under pancuronium administration and SECP was recorded with the scalp (S)-ear (E) monopolar lead, simultaneously. Left column: SECP, SEBP and EEC of patient 10 who remained in a vegetative state. Right column: SECP and SEBP of patient 1 (day 2) who fell into brain death. Lower trace shows EEG at an evoked potential test. EEG became flat on day 3 and the patient died on day 7.
extension and a decrease in amplitude of the latency of the Vth wave immediately before recovery of consciousness (Fig. 2) [9]. Patient 6, who deteriorated into brain death, exhibited a normal appearance of AEBP on the day of resuscitation, delay of latencies on the second day and none of the components of AEBP on the third day. OOR. Even in the patients who regained consciousness, the OOR contained only the R, wave (Table III). In patient 14, who regained consciousness, the R2 wave, with a delayed latency appeared in the test on day 14 after resuscitation and subsequently, the latency was gradually shortened [9]. Both the RI and R2 waves were present even after transition to a vegetative state in some patients, but the latter wave was sharply triphasic (Fig. 6), rather than polyphasic, a common form. Neither the RI nor R2 waves were present in the patients who received diazepam or pentobarbital and in those who became brain dead several days after resuscitation. (3) Spinal Cord SESP, SMR and EEMG (M and H waves). The SESP showed a completely normal pattern even in a vegetative state with markedly disturbed SECP. Examination of conduction changes in the reflex arc of SMR in response to a gradual increase in the stimulating intensity in terms of action potentials of the cauda equina in the subarachnoid space of the lumbar area and H waves of the calf muscle revealed a pattern similar to that in normal individuals [19] (Fig. 7).
243 Table II.
Latency of Vth wave of AEBP. Patients 4 and 6 deteriorated into brain death, patients 9-13 remained in a vegetative state, and patients 14, 15 and 17 recovered consciousness. The standard latency of the Vth wave of AEBP in Japanese adults is 5.65 * 1.0 ms (see Ref. 18). Patient
4 6
9 10
11
12 13 14
No. of days after resuscitation
Latency of Vth wave (ms)
4 la 2” 3 2oa 27a 9” 11” 13” 16a 50” 62” 111” 28
5.3 6.7 5.6 5.6-5.7 5.4 5.5 5.3-5.4 5.5 6.5 6.3 6.2 6.3 5.9 6.3 6.3 6.3 8.4 8.5 6.3 5.4 5.4 6.1
2” 5 9 14 16 29 la 3 1
15 17
‘Under administration of anticonvulsants (diazepam, pentobarbital and/or phenytoin). Table IN.
Appearance of Rt and R, waves of the OOR and simultaneously recorded EEGs. Patients 2 and 5 deteriorated into brain death, patients 8-10 remained in a vegetative state, and patients 14 and 16 recovered consciousness.
Patient
No. of days after resuscitation
OOR RI
2 5 8 9 10 14
16
2 5 81 lga 3a 9 14 16 29 6
EEG R2
+*
+ + +
Isoelectric Isoelectric Spike and fl wave Spike and 6 wave 0 Wave Low voltage 6 wave Low voltage 6 wave Low voltage 0 wave Low voltage 0 wave ci wave
‘Under treatment with anticonvulsants (diazepam, pentobarbital and/or phenytoin). *R2 wave was sharply triphasic.
244 OOR
Rl
fi
h
R2
EEG
1
see
Fig. 6. Sharp triphasic R, wave in vegetated patient 8 (day 81) with increased muscle tone. The lower traces show EEGs recorded monopolarly at that time.
DISCUSSION
In general, disturbance of the CNS following systemic ischemia or anoxia, such as resuscitation from cardiac arrest, extends gradually from the cerebrum, which has weak resistance, to the spinal cord, depending on the severity [20]. For comatose patients immediately after resuscitation from cardiac arrest, objective evaluation of the cerebral function is required. Although the severity of disturbance of the CNS and the patient’s prognosis can be roughly estimated if the interval between cardiac arrest and resuscitation is known, accurate determination of the duration of circulatory arrest is frequently difficult. Furthermore, since many other factors including the situation in which cardiac arrest occurred, underlying diseases and age are also involved, it is difficult to judge the severity and prognosis on the basis of the time factor alone. In this study of evoked potentials, SECP and R2 wave of OOR were used as indicators of cerebral function. The patients who showed all of the early components
245 3
mA
(La-L,)",7;EEMG
