Acta neurol. scandinav. 57, 300-304, 1978 EEG Department, Sectioa of Neurological Sciences, The London Hosspital, London, England
Pathological stimulus-related slow wave arousal responses in the EEG MARTIN S. SCHWARTZ" AND D. F. SCOTT A group of 55 patients were studied who showed in their EEGs prolonged episodes of delta activity, in response to either auditory or tactile stimuli or both. This unusual pattern occurred with states ranging from drowsiness to deep coma but never in alert patients. It was found in a variety of conditions, most commonly following head injury. The EEG phenomenon disappeared within 5 weeks and over half of the patients had a favourable outcome.
The EEGs of patients with impaired consciousness usually show, in response to stimuli, an increase in frequency and a decrease in amplitude of on-going activity. An unusual response characterised by prolonged episodes of delta activity has been demonstrated in patients following head injury (Chatrian et al. 1963, Bickford & Klass 1966). This has also been noted with cerebral haemorrhage or tumour (Evans 1976). The present report concerns a group of patients collected over a 3-year period, all of whom showed this phenomenon. They were studied to determine its aetiology, evolution and prognostic significance. METHOD AND MATERIALS Patients who had a reduced level of consciousness at the time of the EEG were presented with auditory (tapping or calling), tactile (light touch) and painful stimuli to determine their effects. A proportion of these had an abnormal slow wave arousal response. In order to meet our criteria for these responses, they had to show bilateral high voltage delta activity with a frontal distribution. The delta activity could be synchronous o r asynchronotus and was often of a rhythmic nature. To fall within the definition, activity had to persist after the stimulus, last at least 5 seconds, be relatively stereotyped and recur several times in the course of a single EEG. If this phenomenon was noted when the EEG was reported, it was coded and stored on a computer file for subsequent retrieval (Scott 1976). The EEGs were carried out on 8-channel OT 16channel Elema Schonander apparatus with 10120 electrode placement. Clinical and
* Present address: Atkinson Morley's Hospital, Neurology Department, 31, Copse Hill, Wimbledon, S.W. 20, England.
301 Table I . The diagnosis associated with slow wave responses ( n = 55) Head injury Hypotension-hypoxia Cerebral haemorrhage-infarction Uraemia Encephalitis Cerebral tumour Hypoglycaemia Fat embolism
27 9 7 3 5 2 1 1
often EEG follow-up was made t o determine the temporal relationship between the presumed causative condition and the phenomenon.
RESULTS
Fifty-five patients, about 0.5 % of all referrals during the 3-year period, showed the defined slow wave response. Of the total number seen, 7 % had impaired consciousness and might have been expected to demonstrate this abnormal phenomenon. The patients ranged in age from 1 to 81 years; there were 35 males and 20 females. In 27 patients there had been a head injury, in nine it followed hypotensive or hypoxic episodes, the latter had occurred following cardiac arrest or cardiac surgery where hypotensive events had been monitored with the cerebral function monitor during the operation (Schwartz et al. 1973). Seven patients had subarachnoid haemorrhage or cerebral infarction and the remainder included cases of uraemia and encephalitis (Table 1). A total of 122 EEGs was obtained in the 55 patients. Thirty-two patients had follow-up tracings, and in some, serial EEGs were possible. Eighty-seven such recordings had evidence of this phenomenon. None of the patients was alert and orientated at the time of the recording containing the phenomenon. Twenty-four were drowsy, restless or confused. Some were able ‘to carry out simple commands and when attempting to speak could utter a few words, but were unable to maintain a conversation. The EEGs in this group usually showed a disorganised ongoing pattern with a mixture of theta, alpha and beta activity (Fig. 1). The patient’s EEG responses to stimuli were sometimes associated with clinical arousal in terms of greater muscle activity and restlessness as well as attempts to communicate. Increases in heart and respiration rates during the time of arousal were monitored in a few patients. The EEG responses lasted up to 8 minutes. The phenomenon was generally more prominent and was longer in duration when the patient showed the greatest behavioural arousal, as demonstrated by an increase in muscle action potentials (Fig. 2). Twenty-two patients were stuporose and 41 were in light or deep coma
302 at the time of their EEGs. In the latter group there were often localised slow wave abnormalities and the on-going activity was usually irregular and in the theta range. The responses were frequently briefer than in the "drowsy" group, continuing for less than 20 seconds. Auditory stimuli were often sufficient to produce the response.
W/ 1 sec
Figure 1 . Sample of EEG f r o m a 49-year-old w o m a n with a right occipital intracerebral haemorrhage which occurred 4 days previously. A haematoma was drained I day prior to the EEG. T h e patient was in light stupor and showed prolonged episodes of delta activity after stimulation.
wvl 1 SEC
Figzire 2. Samples of EEG f r o m a 23-pear-old patient with a head injury and cerebral contusion 4 days before the EEG. T h e putient was stuporose. T h e on-going activity is an irregular admixture of alpha and theta f o r m s . Following tactile stimulation indicated by a i>erticalbar under the first sample of recording a prolonged burst o f delta activity was seen. This commenced approximately 250 millireconds following the stimulur. 130 seconds later the response was still continuing and there was an increase in muscle action potentials. A t 160 secondr following the stimulation the response was less prominent. I t finally ended 4 m i n following the initial stimulus.
303
In 15 of the 87 recordings, in addition to stimulus-related changes, delta episodes occurred without obvious external stimulation. They were then similar in distribution and form but were usually shorter and there was less associated clinical evidence of increased alertness. The stimulus-related slow wave responses were noted as early as 12 h following the patient’s presentation, lasting for a minimum of 6 and a maximum of 34 days. They were briefer and more difficult to elicit in follow-up recordings. In three patients the phenomena were not seen during the initial EEG. These patients were diagnosed as Herpes Simplex encephalitis, cardiac arrest or having fat emboli. Seven patients died within 7 days of presentation. In 38 of the remaining 48, a clinical follow-up of at least 2 months was possible. At this time 21 were clinically normal or had one or more features of the post-traumatic syndrome. Their most recent follow-up EEG, on average SO months following admission, showed a mild diffuse abnormality consisting of irregular theta activity. Eleven were alert but had distinct neurological deficits such as hemiparesis. Six remained unconscious. One of these died 1%.years later without regaining consciousness. In this case slow wave arousals could only be elicited for the first 15 days of the illness. DISCUSSION
Pathological stimulus-related slow wave responses are an uncommon phenomenon and have received scant attention in the literature. They have been described in a wide variety of intra-cranial conditions, for example head injury (Chatrian et al. 1963, Bickford & Klass 1966) and cerebral haemorrhage and tumour (Evans 1976). In the present study about half of the patients had suffered head injuries but vascular, metabolic, infectious conditions were also causative. The patho-physiological basis of the response is unclear but it has been shown experimentally by Kaada et al. (1967) that stimulation of the reticular formation at different levels of anaesthesia produces different results. A response characterised by a desynchronised low voltage record occurs in light levels, while with deepening anaesthesia slow wave bursts are seen which are sometimes asynchronous. At the deepest levels of anaesthesia no change is noted with stimulation, In man the response tends to disappear with deep stages of coma (Li et al. 1952, Fischgold & Mathis 1959). The stimulus-related responses have some similarity to normal arousal phenomena seen in sleep. Further, as the disorder is never seen in the alert patient, indeed always with lowered levels of consciousness and drowsiness to deep coma, it appears to be due to dysfunction of the ascending reticular formation, a view shared by Gloor (1976). The sensory pathways appeared
304
to be relatively preserved as the patients usually respond behaviourally to stimulation. The phenomena is transitory and occurs commonly in acute and only rarely in slowly progressive conditions (Gloor 1976). The responses are more prominent when the patient is at lower levels of consciousness and become shorter and eventually disappear over a few weeks even if the patient’s level of consciousness is not improved. The eventual outcome of our series of patients with 21 of the 48 survivors showing a good recovery and an additional 11 being alert with only minor neurological deficit indicates that the presence of these changes is not necessarily a poor prognostic sign. ACKNOWLEDGEMENTS We would like to thank the physicians and surgeons of the London Hospital for allowing us to study their patients, also Mrs. Jean Held for indefatigable secretarial help.
REFERENCES Bickford,R.G. & Klass,D.W. (1966): EEG changes with acute and chronic head injury. Head lnjury Conference Proceedings, ed. W. F. Caveness & A. E. Walker. Lippincott, Philadelphia. Chatrian, G. E., White, L. E. & Daly, D. (1963): Electroencephalographic patterns resembling those of sleep in certain comatose states after injuries t o the head. Electroenceph. clin. Neurophysiol. 1.5, 272-280. Evans, B. M. (1976): Patterns of arousal in comatose patients. J. Neurol. Neurosurg. and Psychiat. 39, 392-402. Fischgold, H. (& Mattis, P. (1959): Obnubilations, Coma and Stupors. Etude electroencephalographique. Electroenceph. clin. Neurophysiology, Supplement 11. Gloor, P. (1976): Generalised non paroxysmal abnormalities. A. Remond (ed.). Handbook of electroencephalography and clinical neurophysiology, Vod. 11B. pp. 52-88. Elsevier, Amsterdam. Kaada, B. R., Thomas, F., Alanaes, E. & Wester, K. (1967): EEG synchronisation induced by high frequency midbrain reticular stimulation in anaesthetised cats. Electroenceph. clin. Neurophysiol. 22, 220-230. Li, C.L., Jasper,H. & Henderson,L. (1952): The effect of arousal mechanisms on various forms of abnormality in the electroencephalogram. Electroenceph. clin. Neurophysiol. 4, 512-526. Schwartz, M. S., Colvin, M. P., Prior, P. F., Strunin, L., Simpson, B. R., Weaver, E. J. M. & Scott, D. F. (1973): The cerebral function monitor. Anaesthesia 28, 611-618. Scott, D. F. (1976): Understanding EEG. Duckworth, London. Received December 4, 1977, accepted February 21, 1978
Dr. D . F . Scott EEG Department The London Hospital Whitechapel London E.1.1BB England
Pathological stimulus-related slow wave arousal responses in the EEG MARTIN S. SCHWARTZ" AND D. F. SCOTT A group of 55 patients were studied who showed in their EEGs prolonged episodes of delta activity, in response to either auditory or tactile stimuli or both. This unusual pattern occurred with states ranging from drowsiness to deep coma but never in alert patients. It was found in a variety of conditions, most commonly following head injury. The EEG phenomenon disappeared within 5 weeks and over half of the patients had a favourable outcome.
The EEGs of patients with impaired consciousness usually show, in response to stimuli, an increase in frequency and a decrease in amplitude of on-going activity. An unusual response characterised by prolonged episodes of delta activity has been demonstrated in patients following head injury (Chatrian et al. 1963, Bickford & Klass 1966). This has also been noted with cerebral haemorrhage or tumour (Evans 1976). The present report concerns a group of patients collected over a 3-year period, all of whom showed this phenomenon. They were studied to determine its aetiology, evolution and prognostic significance. METHOD AND MATERIALS Patients who had a reduced level of consciousness at the time of the EEG were presented with auditory (tapping or calling), tactile (light touch) and painful stimuli to determine their effects. A proportion of these had an abnormal slow wave arousal response. In order to meet our criteria for these responses, they had to show bilateral high voltage delta activity with a frontal distribution. The delta activity could be synchronous o r asynchronotus and was often of a rhythmic nature. To fall within the definition, activity had to persist after the stimulus, last at least 5 seconds, be relatively stereotyped and recur several times in the course of a single EEG. If this phenomenon was noted when the EEG was reported, it was coded and stored on a computer file for subsequent retrieval (Scott 1976). The EEGs were carried out on 8-channel OT 16channel Elema Schonander apparatus with 10120 electrode placement. Clinical and
* Present address: Atkinson Morley's Hospital, Neurology Department, 31, Copse Hill, Wimbledon, S.W. 20, England.
301 Table I . The diagnosis associated with slow wave responses ( n = 55) Head injury Hypotension-hypoxia Cerebral haemorrhage-infarction Uraemia Encephalitis Cerebral tumour Hypoglycaemia Fat embolism
27 9 7 3 5 2 1 1
often EEG follow-up was made t o determine the temporal relationship between the presumed causative condition and the phenomenon.
RESULTS
Fifty-five patients, about 0.5 % of all referrals during the 3-year period, showed the defined slow wave response. Of the total number seen, 7 % had impaired consciousness and might have been expected to demonstrate this abnormal phenomenon. The patients ranged in age from 1 to 81 years; there were 35 males and 20 females. In 27 patients there had been a head injury, in nine it followed hypotensive or hypoxic episodes, the latter had occurred following cardiac arrest or cardiac surgery where hypotensive events had been monitored with the cerebral function monitor during the operation (Schwartz et al. 1973). Seven patients had subarachnoid haemorrhage or cerebral infarction and the remainder included cases of uraemia and encephalitis (Table 1). A total of 122 EEGs was obtained in the 55 patients. Thirty-two patients had follow-up tracings, and in some, serial EEGs were possible. Eighty-seven such recordings had evidence of this phenomenon. None of the patients was alert and orientated at the time of the recording containing the phenomenon. Twenty-four were drowsy, restless or confused. Some were able ‘to carry out simple commands and when attempting to speak could utter a few words, but were unable to maintain a conversation. The EEGs in this group usually showed a disorganised ongoing pattern with a mixture of theta, alpha and beta activity (Fig. 1). The patient’s EEG responses to stimuli were sometimes associated with clinical arousal in terms of greater muscle activity and restlessness as well as attempts to communicate. Increases in heart and respiration rates during the time of arousal were monitored in a few patients. The EEG responses lasted up to 8 minutes. The phenomenon was generally more prominent and was longer in duration when the patient showed the greatest behavioural arousal, as demonstrated by an increase in muscle action potentials (Fig. 2). Twenty-two patients were stuporose and 41 were in light or deep coma
302 at the time of their EEGs. In the latter group there were often localised slow wave abnormalities and the on-going activity was usually irregular and in the theta range. The responses were frequently briefer than in the "drowsy" group, continuing for less than 20 seconds. Auditory stimuli were often sufficient to produce the response.
W/ 1 sec
Figure 1 . Sample of EEG f r o m a 49-year-old w o m a n with a right occipital intracerebral haemorrhage which occurred 4 days previously. A haematoma was drained I day prior to the EEG. T h e patient was in light stupor and showed prolonged episodes of delta activity after stimulation.
wvl 1 SEC
Figzire 2. Samples of EEG f r o m a 23-pear-old patient with a head injury and cerebral contusion 4 days before the EEG. T h e putient was stuporose. T h e on-going activity is an irregular admixture of alpha and theta f o r m s . Following tactile stimulation indicated by a i>erticalbar under the first sample of recording a prolonged burst o f delta activity was seen. This commenced approximately 250 millireconds following the stimulur. 130 seconds later the response was still continuing and there was an increase in muscle action potentials. A t 160 secondr following the stimulation the response was less prominent. I t finally ended 4 m i n following the initial stimulus.
303
In 15 of the 87 recordings, in addition to stimulus-related changes, delta episodes occurred without obvious external stimulation. They were then similar in distribution and form but were usually shorter and there was less associated clinical evidence of increased alertness. The stimulus-related slow wave responses were noted as early as 12 h following the patient’s presentation, lasting for a minimum of 6 and a maximum of 34 days. They were briefer and more difficult to elicit in follow-up recordings. In three patients the phenomena were not seen during the initial EEG. These patients were diagnosed as Herpes Simplex encephalitis, cardiac arrest or having fat emboli. Seven patients died within 7 days of presentation. In 38 of the remaining 48, a clinical follow-up of at least 2 months was possible. At this time 21 were clinically normal or had one or more features of the post-traumatic syndrome. Their most recent follow-up EEG, on average SO months following admission, showed a mild diffuse abnormality consisting of irregular theta activity. Eleven were alert but had distinct neurological deficits such as hemiparesis. Six remained unconscious. One of these died 1%.years later without regaining consciousness. In this case slow wave arousals could only be elicited for the first 15 days of the illness. DISCUSSION
Pathological stimulus-related slow wave responses are an uncommon phenomenon and have received scant attention in the literature. They have been described in a wide variety of intra-cranial conditions, for example head injury (Chatrian et al. 1963, Bickford & Klass 1966) and cerebral haemorrhage and tumour (Evans 1976). In the present study about half of the patients had suffered head injuries but vascular, metabolic, infectious conditions were also causative. The patho-physiological basis of the response is unclear but it has been shown experimentally by Kaada et al. (1967) that stimulation of the reticular formation at different levels of anaesthesia produces different results. A response characterised by a desynchronised low voltage record occurs in light levels, while with deepening anaesthesia slow wave bursts are seen which are sometimes asynchronous. At the deepest levels of anaesthesia no change is noted with stimulation, In man the response tends to disappear with deep stages of coma (Li et al. 1952, Fischgold & Mathis 1959). The stimulus-related responses have some similarity to normal arousal phenomena seen in sleep. Further, as the disorder is never seen in the alert patient, indeed always with lowered levels of consciousness and drowsiness to deep coma, it appears to be due to dysfunction of the ascending reticular formation, a view shared by Gloor (1976). The sensory pathways appeared
304
to be relatively preserved as the patients usually respond behaviourally to stimulation. The phenomena is transitory and occurs commonly in acute and only rarely in slowly progressive conditions (Gloor 1976). The responses are more prominent when the patient is at lower levels of consciousness and become shorter and eventually disappear over a few weeks even if the patient’s level of consciousness is not improved. The eventual outcome of our series of patients with 21 of the 48 survivors showing a good recovery and an additional 11 being alert with only minor neurological deficit indicates that the presence of these changes is not necessarily a poor prognostic sign. ACKNOWLEDGEMENTS We would like to thank the physicians and surgeons of the London Hospital for allowing us to study their patients, also Mrs. Jean Held for indefatigable secretarial help.
REFERENCES Bickford,R.G. & Klass,D.W. (1966): EEG changes with acute and chronic head injury. Head lnjury Conference Proceedings, ed. W. F. Caveness & A. E. Walker. Lippincott, Philadelphia. Chatrian, G. E., White, L. E. & Daly, D. (1963): Electroencephalographic patterns resembling those of sleep in certain comatose states after injuries t o the head. Electroenceph. clin. Neurophysiol. 1.5, 272-280. Evans, B. M. (1976): Patterns of arousal in comatose patients. J. Neurol. Neurosurg. and Psychiat. 39, 392-402. Fischgold, H. (& Mattis, P. (1959): Obnubilations, Coma and Stupors. Etude electroencephalographique. Electroenceph. clin. Neurophysiology, Supplement 11. Gloor, P. (1976): Generalised non paroxysmal abnormalities. A. Remond (ed.). Handbook of electroencephalography and clinical neurophysiology, Vod. 11B. pp. 52-88. Elsevier, Amsterdam. Kaada, B. R., Thomas, F., Alanaes, E. & Wester, K. (1967): EEG synchronisation induced by high frequency midbrain reticular stimulation in anaesthetised cats. Electroenceph. clin. Neurophysiol. 22, 220-230. Li, C.L., Jasper,H. & Henderson,L. (1952): The effect of arousal mechanisms on various forms of abnormality in the electroencephalogram. Electroenceph. clin. Neurophysiol. 4, 512-526. Schwartz, M. S., Colvin, M. P., Prior, P. F., Strunin, L., Simpson, B. R., Weaver, E. J. M. & Scott, D. F. (1973): The cerebral function monitor. Anaesthesia 28, 611-618. Scott, D. F. (1976): Understanding EEG. Duckworth, London. Received December 4, 1977, accepted February 21, 1978
Dr. D . F . Scott EEG Department The London Hospital Whitechapel London E.1.1BB England
