Postanoxic alpha (theta) coma: a reappraisal of its prognostic significance

Clinical Neurophysiology 111 (2000) 297±304

Clinical insight

www.elsevier.com/locate/clinph

Postanoxic alpha (theta) coma: a reappraisal of its prognostic signi®cance Magdalena Berkhoff, Filippo Donati, Claudio Bassetti* Department of Neurology, University Hospital, Inselspital, 3010Berne, Switzerland Accepted 14 September 1999

Abstract Objectives: To appraise the controversial prognostic signi®cance of postanoxic alpha or theta coma (ATC). Methods: We prospectively assessed 14 comatose patients with ATC after cardiac arrest by means of a protocol which included repeated clinical examinations, EEG, and median somatosensory evoked potentials (SEP). Good outcome was de®ned by the reappearance of cognition (Glasgow outcome scale 3±5) at any time during the 1 year follow-up. Results: Nine of 14 patients had a monotonous, frontally accentuated and areactive alpha (theta) EEG activity (complete ATC). In these patients ATC was recorded a mean of 47 h after resuscitation, the mean Glasgow coma scale (GCS) was 4 at 48 h, and early cortical SEPs were altered or absent in 5 of 7 patients. All nine patients died. In ®ve of 14 patients the alpha (theta) EEG activity was either not monotonous, partially reactive or posteriorly dominant (incomplete ATC). In these patients ATC was recorded a mean of 43 h after resuscitation, the mean GCS was 8 at 48 h, and early cortical SEP were normal in 4 of 5 patients. Three of 5 patients regained cognition, two of them remained however dependent in activities of everyday life. Conclusions: This study and a review 283 cases of postanoxic ATC reported in the literature suggest the existence of incomplete and complete variants of postanoxic ATC. Whereas complete ATC is invariably associated with a poor outcome, full recovery is possible in patients with incomplete ATC. The combination of EEG, clinical, and SEP ®ndings improves the prognostic accuracy of postanoxic ATC. q 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Alpha-theta coma; Postanoxic coma; Cardiac arrest; EEG; Somatosensory evoked potentials; Outcome; Prognosis

1. Introduction Alpha coma refers to EEG patterns predominating in the alpha frequency band recorded in comatose patients. Single reports of postanoxic alpha coma after cardiac arrest ®rst appeared in the late, 1960s and early, 1970s in patients with a poor outcome (Hockaday et al., 1965; Lindgren et al., 1968; Binnie et al., 1970; Brierley et al., 1971; Lemmi et al., 1973). Vignaendra et al. (1974) reported a ®rst series of eight patients with postanoxic alpha coma, none of whom survived (Vignaendra et al., 1974). Likewise, all 5 patients with postanoxic alpha coma described by Westmoreland et al. (1975) died within 21 days. Following a few studies con®rming these early observations (MoÈller et al., 1978; Alving et al., 1979), single reports of patients with postanoxic alpha coma and recovery of cognition appeared in the literature (Chokroverty, 1975; Grindal et al., 1977; Gottschalk et al., 1978; Soerensen et al., 1978; Iragui and McCutchen, 1983). More recently, survival with regained * Corresponding author. Tel.: 141-31-632-3066; fax: 141-31-6329679. E-mail address: [email protected] (C. Bassetti)

consciousness in coma after cardiac arrest was reported in as many as six (12%) of 49 patients by Austin et al. (1988). Synek et al. ®rst described a theta variant of alpha coma (theta coma) emphasizing its association with a poor outcome (Synek and Synek, 1984, 1987; Janati et al., 1986b). The transition from theta coma to alpha coma and vice versa (Synek and Synek, 1988; Bortone et al., 1994), and the coexistence of both EEG patterns in a few patients with postanoxic coma (Janati et al., 1986a) support the hypothesis of a common pathophysiologic mechanism (Synek and Synek, 1984). As for alpha coma, the association of theta coma with a poor outcome assumed in early reports was recently questioned. Indeed, Young et al. reported survival with regained consciousness in 6 (13%) of 46 patients with alpha coma, theta coma, or alpha-thetacoma (Young et al., 1994). At present time the value of postanoxic alpha or theta coma (ATC) in predicting outcome remains therefore controversial. The aim of the present study was to reassess the prognostic signi®cance of postanoxic ATC. For this purpose we performed (1) a prospective analysis of clinical, EEG, and evoked potential ®ndings of 14 patients with ATC

1388-2457/00/$ - see front matter q 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 1388-245 7(99)00246-1

CLINPH 98668

298

M. Berkhoff et al. / Clinical Neurophysiology 111 (2000) 297±304

after cardiac arrest and (2) a review of the pertinent literature. 2. Patients and methods The EEG recordings of 60 previously reported and prospectively assessed patients in coma after cardiac arrest were reviewed for the presence of an EEG pattern predominating in the alpha or theta frequency band (ATC) (Bassetti et al., 1996). Patients with normally reactive and normally (occipitally) distributed predominant alpha, alpha-theta or theta EEG activities were not considered. Patients with associated factors likely to affect the EEG such as metabolic disorders, head trauma or drug-induced coma were also excluded. Clinical examinations were repeated 6±12 h, 24 h, 48 h and 7 days after resuscitation. Clinical assessment included the Glasgow coma scale (GCS) (Mullie et al., 1988; Sacco et al., 1990) and the examination of brain-stem re¯exes (corneal, pupillar and oculocephalic re¯exes) (Snyder et al., 1981; Edgren et al., 1994). Electroencephalogram (EEG) was recorded in all patients with an 8±16 channel machine. Scalp electrodes were placed according to the International 10±20 system. The ®rst EEG was recorded for at least 30 min within 5±72 h following cardiac arrest. Reactivity to acoustic, painful and photic stimuli was tested in all cases. In each patient, one to 5 EEGs were recorded. In one patient a long-term EEGmonitoring was performed. Somatosensory evoked potentials (SEP) were obtained by electrical stimulation of the median nerve at the wrist with square pulses of 0.2 ms duration and an intensity of 4 mA above motor threshold of the thenar muscle delivered at 3±5 Hz. Bipolar recordings were made with a Medelec sensor device and bandpass ®lters from 3 Hz to 3 kHz with the recording needles placed according to the 10±20 nomenclature: (a) ipsilateral Erb's point to Fz; (b) spinous process of C7 to Fz; (c) contralateral hand®eld (P3/P4) to ipsilateral ear (A1/A2); (d) P3/P4 to Fz. Two samples each of 512 to 1024 sweeps were averaged on either side. The latencies of the N13 1, P15 (some authors label this subcortical SEP component P14), N20 and P25 responses were determined and the central conduction time (CCT) and the P25/N20 amplitude ratio were calculated (as shown in Fig. 1). The P25/N20 amplitude ratio and the CCT were considered normal if within 2.5 SD of healthy controls. Results of cortical SEP were classi®ed as (a) normal (normal N20/P25 amplitude 1 normal central conduction time), (b) abnormal (N20/P25 amplitude reduction .50% or CCT . 7:2 ms uni- or bilaterally), or (c) absent (no N20/P25 bilaterally). Outcome was de®ned as the best cerebral performance achieved at any time over a follow-up period of 1 year. 1 Since cervical recordings were performed with a frontal reference spurious contamination of N13 by P14 cannot be excluded in our patients (as visible in Fig. 1).

Fig. 1. Scoring of median somatosensory evoked potentials.

Good outcome was de®ned by the reappearance of cognition corresponding to a Glasgow outcome scale (GOS) (Jennett and Bond, 1975) score of 3±5. Bad outcome was de®ned by a persistent vegetative state (Zeman, 1997) or death (GOS 1±2). 3. Results An EEG pattern predominating in the alpha or theta frequency band (ATC) was found in 14 (23%) of 60 consecutive patients with postanoxic coma after cardiac arrest. There were 13 men and one woman with a mean age of 58 years (range: 24±83). Clinical and electrophysiological ®ndings are summarized in Table 1. 3.1. EEG ATC was recorded 5 h after resuscitation in one patient and between 24 and 72 h in the remaining 13 patients. In 7 patients ATC consisted of an alpha activity (8±12 Hz) that was monotonous (little changes in amplitude), continuous (little changes in frequency), frontally distributed, and areactive to external stimuli (Fig. 2). In two other patients there was a monotonous, continuous, areactive frontally accentuated theta-activity (4±7 Hz). We considered the EEG patterns in these nine patients as representative of complete ATC. The remaining 5 patients showed a predo-

M. Berkhoff et al. / Clinical Neurophysiology 111 (2000) 297±304 Table 1 Clinical and electrophysiological ®ndings in 14 patients with postanoxic alpha (theta) coma (ATC)

299

re¯exes were present in two out of 5 patients at 6±12 h, and in 4 out of 5 patients at 48 h.

Complete a ATC

Incomplete a ATC

3.3. SEP ®ndings

Number of patients Male:female Mean age in years (range)

9 8:1 58 (24±83)

5 5:0 57 (50±72)

Glasgow coma Score At 24 h, mean (range) At 48 h

4 (3±7) 4 (3±6)

5 (3±9) 8 (6±13)

EEG Interval to recording (range)

47 h (24±72)

43 h (5±72)

In seven patients SEP recordings were carried out within 2 h from the EEG recording showing ATC, and in the remaining ®ve cases within 14±48 h. Somatosensory evoked potentials were recorded in 7 out of 9 patients with complete ATC and in all 5 patients with incomplete ATC. In the seven patients with complete ATC early cortical SEP were absent in four, abnormal in one, and normal in two. In the 5 patients with incomplete ATC early cortical SEP were normal in four and absent in one. In both groups, all patients with absent cortical SEP died.

53 h (24±96) 4 of 7

58 h (34±96) 3 of 5

2 of 7

4 of 5

0

3

SEP Interval to recording (range) SEP recorded simultaneously to EEG Normal SEP Good outcome a

a

See text for de®nitions.

minant alpha or theta activity that was either clearly not monotonous (Fig. 3), hyporeactive, or areactive but posteriorly distributed (incomplete ATC). In 8 patients with poor outcome serial EEG recordings revealed in both complete (n ˆ 6) and incomplete (n ˆ 2) ATC the progressive ¯attening of the EEG activity with appearance of intermittent isoelectric pauses preceding death. This evolution is well illustrated by the following case report. In a 24 year old man with coma after cardiac arrest the ®rst EEG recorded 8 h after resuscitation shows a diffuse, areactive, low voltage 14±16 Hz activity. At 72 h the EEG is unchanged but after painful an irregular alpha activity appears particularly over the fronto-central areas. A long-term EEG recording is started at 79 h. At 81 h the activity consists of an anteriorly dominant 14±16 Hz betaactivity, areactive to external stimuli. At 92 h a 6 Hz anteriorly dominant, areactive theta activity, alternating with episodes of bilateral suppression of brain activity appear (theta coma). These periods last initially for about 4 s, but duration progressively increases to 14±20 s at 100 h after resuscitation. Six days after cardiac arrest the patient died without awakening from coma. 3.2. Clinical ®ndings All 9 patients with complete ATC had at 6±12 h a GCS of 3. At 24 h the mean GCS was 4 (range: 3±7) without improvement at 48 h. Brain-stem re¯exes were preserved in 5 out of 9 patients at 6±12 h and in 3 out of 7 patients at 48 h. In patients with incomplete ATC the mean GCS at 6±12 h was 5 (range: 3±5). At 24 h the mean GCS was 5 (range: 3± 9) and improved to 8 (range: 6±13) at 48 h. Brain-stem

3.4. Outcome All patients with complete ATC died. Eight patients died within the ®rst 10 days and one after a survival of 8 months in a persistent vegetative state. Three of ®ve patients with incomplete ATC survived and regained cognition. Two of them remained dependent in activities of everyday life (GOS 3). 4. Discussion This is the ®rst study of a series of patients with postanoxic alpha-theta coma (ATC) in which clinical ®ndings, SEP, and outcome were prospectively assessed. The main ®nding of the present analysis is the identi®cation of two forms of postanoxic ATC. Complete ATC ± with a monotonous, continuous, frontally distributed, areactive EEG activity was typically associated with abnormal or absent cortical SEP, deep coma with little improvement in the ®rst 48 h after cardiac arrest, and poor outcome. Conversely, in patients with incomplete ATC ± with not monotonous, posteriorly accentuated or partially reactive EEG tracings ± early cortical SEP were mostly normal, and good outcome with recovery of cognition was observed. Our suggestion of the existence of complete and incomplete ATC presumably explains contradictory results previously reported concerning prevalence and prognosis of ATC. Some authors required for the diagnosis of alpha coma only the presence of a predominant alpha EEG activity, regardless of EEG response to external stimuli, presence or absence of activities in other frequency bands, and distribution of the EEG activity (Chokroverty, 1975; Soerensen et al., 1978; Austin et al., 1988). As a consequence, ATC was reported to be present in up to 26% of patients with postanoxic coma and to lack prognostic signi®cance (Grindal et al., 1977; Soerensen et al., 1978; Austin et al., 1988; Young et al., 1994). In a non-selected review of the literature we found that 33 (12%) of 283 patients with postanoxic ATC survived, 16 of whom with only minimal or no residual cognitive de®cits (Table 2, Lindgren et al., 1968; Brier-

300

M. Berkhoff et al. / Clinical Neurophysiology 111 (2000) 297±304

Fig. 2. Complete alpha coma after cardiac arrest. (a) The EEG of an 83 year old man recorded 48 h after cardiac arrest shows a complete alpha coma with a low-amplitude, areactive alpha activity that is maximal over the frontal regions. Arrow: passive eye opening, double arrow: passive eye closure. (b) The SEP recorded within a few min after the EEG examination reveal absent early cortical responses. The patient died the following day.

ley et al., 1971; Vignaendra et al., 1974; Chokroverty, 1975; Grindal and Suter, 1975; Westmoreland et al., 1975; Grindal et al., 1977; Snyder et al., 1977; Gottschalk et al., 1978; MoÈller, 1978; MoÈller et al., 1978; Soerensen et al., 1978; Alving et al., 1979; Bermejo Pareja et al., 1979; Heraut et al., 1980; Obeso et al., 1980; Iragui and McCutchen, 1983; Synek and Synek, 1984; Scott and Sumra, 1985; Janati et al., 1986b; Ganji et al., 1987; Scollo-Lavizzari and Bassetti, 1987; Synek and Synek, 1987; Synek and Synek, 1988; Austin et al., 1988; Bassetti and Karbowski, 1990; Uldry et al., 1991; Bortone et al., 1994; Young et al., 1994; Kaplan et al., 1999). However, if a more restrictive de®nition of ATC is used ± such as those of complete ATC suggested in this study and for ATC in some of the previous studies (Binnie et al., 1970; MoÈller, 1978; Synek, 1990) ± the frequency of ATC in postanoxic coma does not exceed 10±15% (MoÈller et al., 1978; Soerensen et al., 1978; Alving

et al., 1979; Bassetti et al., 1996) and ± more importantly ± outcome is almost invariably poor. Indeed, a favorable evolution was reported in only two adults with complete ATC. Iragui and McCutchen reported a 64 year old patient in whom ATC was recorded 2 h after cardiac arrest and was followed by an almost complete clinical recovery (Iragui and McCutchen, 1983). SoÈrensen et al. reported a 34 year old patient with complete ATC recorded 5 days after cardiac arrest, in whom only minimal residual de®cits (memory impairment) were noted after awakening (Soerensen et al., 1978). All other patients reported in the literature with good outcome in whom the EEG ®ndings were reported in detail had an incomplete variant of ATC (Table 2) or were below the age of 18 years (Soerensen et al., 1978; Collins and Chatrian, 1980). It is noteworthy, that neither a frontal distribution nor an absent reactivity of the EEG activity appear to be suf®cient, when considered in isolation, to

M. Berkhoff et al. / Clinical Neurophysiology 111 (2000) 297±304

301

Fig. 3. Incomplete alpha coma after cardiac arrest. (a) The EEG of a 47 year old man recorded 34 h after cardiac arrest shows an incomplete alpha coma with diffusely distributed, not monotonous, and areactive alpha activity. (b) The SEP recorded the same day are normal. The patient survived with minimal cognitive de®cits.

predict a poor outcome in postanoxic ATC (Grindal et al., 1977; Austin et al., 1988; Bassetti and Karbowski, 1990; Young et al., 1994). In conclusion, ATC represents an (almost) infaust prognostic sign in postanoxic coma as long as (1) the alpha EEG

activity if frontally distributed; and (2) continuous (little changes in frequency), monotonous (little changes in amplitude) and areactive to external stimuli; (3) the patient is an adult; and (4) the recording is performed . 12±24 h after cardiac arrest.

302

M. Berkhoff et al. / Clinical Neurophysiology 111 (2000) 297±304

Table 2 Postanoxic alpha (theta) coma (literature review a) Author, year

Survivors/no. of patients

Lindgren et al. (1968) Brierley et al. (1971) Vignaendra et al. (1974) Chokroverty (1975)

0/1 0/1 0/8 2/12

Westmoreland et al. (1975) Grindal et al. (1977) c

0/5 3/14

Snyder et al. (1977) Soerensen et al. (1978)

1/2 3/13

Gottschalk et al. (1978) MoÈller et al. (1978) c Bermejo Pareja et al. (1979) Heraut et al. (1980) Obeso et al. (1980) Iragui and McCutchen (1983) Scott and Sumra (1985) Janati et al. (1986) Ganji et al. (1987) Synek and Synek (1987) c Scollo-Lavizzari and Bassetti (1987) Austin et al. (1988)

1/5 1/35 0/6 0/2 0/2 1/1 0/1 0/1 0/2 0/11 1/5

Bassetti and Karbowski (1990)

2/12

Uldry et al. (1991) Bortone et al. (1994) Young et al. (1994)

3/16 0/1 6/43

Kaplan et al. (1999) Berkhoff et al. (this study)

1/21 3/14

Total

33/283 (12%)

5/49

Comments

b

One patient with mild memory de®cits and diffuse AC 1 day after CA; 1 patient with unknown severity of de®cits One patient with normal mentation but unknown AC characteristics; 2 patients with severe cognitive de®cits Mild cognitive de®cits but unknown AC characteristics One patient with severe cognitive de®cits; 1 patient (16 years old) with complete recovery and frontal/areactive AC 1 day after CA; 1 patient with minimal cognitive de®cits and frontal/areactive AC 5 days after CA Cognitive de®cits? Severe cognitive de®cits (dementia)

Mild cognitive de®cits and frontal/areactive AC 2 h after CA TC TC Severe cognitive de®cits One patient with no gross cognitive de®cits; four patients with cognitive de®cits (severity?); AC characteristics unknown in single patients One patient with mild cognitive de®cits and frontal/reactive AC 2 days after CA; 1 patient with severe cognitive de®cits One patient with complete recovery but unknown AC characteristics Transition from TC to AC Six patients with `conscious awareness' but unknown ATC characteristics in single patients Severe cognitive de®cits AC, TC, 2 patients with severe cognitive de®cits; 1 patient with mild cognitive de®cits and diffuse AC

a

Only adult cases with cardiopulmonary arrest were considered. AC, alpha coma; TC, theta coma; ATC, alpha-theta coma. c These authors reported one or more cases in separate publications (Grindal and Suter, 1975; Grindal et al., 1977; MoÈller, 1978; MoÈller et al., 1978; Alving et al., 1979; Synek and Synek, 1984, 1987, 1988). b

The pathogenesis of ATC in postanoxic coma is unknown. In dogs, posthypoxic alpha-like activity appears ®rst and with maximum amplitudes in the amygdaloid nuclei (Gurvitch et al., 1984). In humans, amygdala, hypothalamus, brain-stem and basal forebrain are relatively preserved after hypoxic injury (Kinney and Samuels, 1994). Conversely, diffuse laminar necrosis of the neocortex or selective necrosis of the thalamus is common (Fujioka et al., 1994; Kinney and Samuels, 1994). These autoptic ®ndings have also been documented in a few patients with ATC (Brierley et al., 1971; Chokroverty, 1975; Grindal et al., 1977; Heraut et al., 1980) and suggest that, once thalamocortical neurons are damaged, the amygdala or other subcortical structures may function as `last pacemakers' of electric

brain activity. This hypothesis could explain the monotonous, synchronized, frontally accentuated, areactive EEG activity and poor outcome of complete ATC. Dysfunction of thalamo-cortical pathways in patients with complete ATC is further supported by three lines of evidence. First, complete ATC would exclude the presence of sleep activity, since sleep activity requires intact thalamo-cortical pathways. Accordingly, sleep activity was absent in previously reported patients with complete ATC (MoÈller, 1978; Synek and Synek, 1984). Second, ATC may coexist with burstsuppression pattern (Janati et al., 1986a), which results from a disconnection of cortical neurons from cortical gray matter (Henry and Scoville, 1952). Third, changes of early cortical SEP ± which re¯ect the extent of anoxic

M. Berkhoff et al. / Clinical Neurophysiology 111 (2000) 297±304

303

Table 3 Alpha (theta) activities in adult patients with anoxic brain damage and decreased level of consciousness

EEG(alpha activity) SEP/(early cortical responses) Clinical examination Outcome

EEG Grade I a

Incomplete ATC

Complete ATC

Occipital, not monotonous, reactive Normal Somnolence or sopor, normal brain-stem re¯exes Usually good

Occipital or diffuse, not monotonous, hypo-areactive

Frontal, monotonous, areactive

Usually normal Coma, usually normal brainstem re¯exes

Usually altered or absent Deep coma, usually absent brain-stem re¯exes

Often poor, recovery possible

Invariably poor, unless recorded , 12±24 h after cardiac arrest

a

According to a classi®cation of EEG patterns in postanoxic coma into 5 grades (MoÈller et al., 1978; Scollo-Lavizzari and Bassetti, 1987; Bassetti and Karbowski, 1990; Synek, 1990). SEP, median somatosensory evoked potentials.

damage of thalamo-cortical pathways (Trojaborg and JoÈrgensen, 1973; Walser et al., 1986; Bassetti et al., 1996) were found in most patients with complete ATC (Westmoreland et al., 1975; Obeso et al., 1980; Synek and Synek, 1984, 1987; Young et al., 1994). The occasional preservation of cortical SEP in complete ATC, as reported by Brierley et al. and seen in two of our patients, may be related to a preferential involvement of neurons in the paramedian thalamus compared to the somatosensory relay nuclei of the thalamus (Brierley et al., 1971). It is at present time unclear whether incomplete ATC always corresponds to a variant of complete alpha coma. Our clinical and SEP observations demonstrate that incomplete ATC re¯ects a less severe degree of anoxic brain damage than complete ATC. Also, an evolution from incomplete to complete ATC has been described, whereas a transition from complete to incomplete ATC was never reported. These ®ndings suggest that incomplete ATC may ± in some patients ± represent a residuum of physiologic alpha EEG activity. Based on our data and review of the literature we suggest to differentiate between three different types of alpha (theta) activities in patients with anoxic brain damage and decreased level of consciousness (Table 3): 1. EEG grade I, which usually carries a good prognosis; 2. Incomplete ATC, in which outcome is often but not always poor; 3. Complete ATC, which almost invariably heralds a poor outcome. The differentiation between (1) and (2), and particularly between (2) and (3) may not be always trivial. In dubious cases the clinician should take into consideration not only the EEG but also clinical and, when available, SEP ®ndings. The combination of the 3 parameters enhances the prognostic accuracy in postanoxic coma (Bassetti et al., 1996). Absent cortical SEPs remain, at the present time, the single most accurate predictor of poor outcome in postanoxic coma (Zandbergen et al., 1998).

References Alving J, MoÈller M, Sindrup E, Nielson BL. Alpha pattern coma following cerebral anoxia. Electroenceph clin Neurophysiol 1979;47:95± 101. Austin E, Wilkus RJ, Longstreth WT. Etiology and prognosis of alpha coma. Neurology 1988;33:773±777. Bassetti C, Karbowski K. Prognostischer Wert der Elektroenzephalographie bei nicht-traumatischen Komata. Schweiz Med Wschr 1990;120:1425± 1434. Bassetti C, Bomio F, Mathis J, Hess CW. Prognosis of coma after cardiac arrest: A prospective study of 60 patients. J Neurol Neurosurg Psychiatry 1996;61:610±615. Bermejo Pareja F, Martinez Martin M, Alvarez Tejerina A. `Coma alfa'. Arch Neurobiol 1979;42:243±262. Binnie CD, Prior PF, Lloyd DSL, Scotts DF, Margerison JH. Electroencephalographic prediction of fatal anoxic brain damage after resuscitation from cardiac arrest. Br Med J 1970;4:265±268. Bortone E, Bettoni L, Giorgi C, Trabattoni GR, Macia D. Post-anoxic theta and alpha pattern coma. Clin Electrophysiol 1994;25:156±159. Brierley JB, Graham DI, Adams JH. Neocortical death after cardiac arrest. Lancet 1971;2:560±565. Chokroverty S. `Alpha-like' rhythms in electroencephalograms in coma after cardiac arrest. Neurology 1975;25:655±663. Collins T, Chatrian GE. EEG rhythm of alpha frequency in a 22-month-old child after strangulation. Neurology 1980;30:1316±1319. Edgren E, Hedstrand U, Kelsey S, Sutton-Tyrell K, Safar P. and BRCT I Study Group. Assessment of neurological prognosis in comatose survivors of cardiac arrest. Lancet 1994;343:1055±1059. Fujioka M, Okuchi K, Sakaki T, Hiramatsu KI, Miyamoto S, Iwasaki S. Speci®c changes in human brain following reperfusion after cardiac arrest. Stroke 1994;25:2091±2095. Ganji S, Peters G, Frazier E. Alpha-coma: clinical and evoked potential studies. Clin Electroenceph 1987;18:103±113. Gottschalk PG, Hansotia P, Berendes J. Alpha coma: report of eleven cases (abstract). Electroenceph clin Neurophysiol 1978;44:122. Grindal AB, Suter C. `Alpha-pattern coma' in high voltage electrical injury. Electroenceph clin Neurophysiol 1975;38:521±526. Grindal AB, Suter C, Martinez AJ. Alpha-pattern coma: 24 cases with 9 survivors. Ann Neurol 1977;1:371±377. Gurvitch AM, Zarhetsky YV, Trush VD, Zonov VM. Experimental data on the nature of postresuscitation alpha frequency activity. Electroenceph clin Neurophysiol 1984;58:426±437. Henry CE, Scoville WB. Suppression-burst activity from isolated cerebral cortex in man. Electroenceph clin Neurophysiol 1952;4:1±22. Heraut LA, Lombard C, Cathala HP. Coma profond a frequence alpha apreÁs

304

M. Berkhoff et al. / Clinical Neurophysiology 111 (2000) 297±304

pendaison non suivie de mort immeÂdiate. Rev EEG Neurophysiol 1980;10:21±32. Hockaday JM, Potts F, Epstein E, Bonazzi A, Schwab R. Electroencephalographic changes in acute cerebral anoxia from cardiac or respiratory arrest. Electroenceph clin Neurophysiol 1965;18:575±586. Iragui V, McCutchen CB. Physiologic and prognostic signi®cance of `alpha coma'. J Neurol Neurosurg Psychiatry 1983;46:632±638. Janati A, Husain MM, Moore DB, Adametz JR. Suppression-burst pattern associated with generalized epileptiform discharges and alpha-theta pattern coma. Clin Electroenceph 1986a;17:82±88. Janati A, Husain MM, Moore DB, et al. Coexistence of ectopic rhythms and periodic EEG pattern in anoxic encephalopathy. Clin Electroenceph 1986b;17:187±194. Jennett B, Bond M. Assessment of outcome after severe brain damage. A practical scale. Lancet 1975;1:480±484. Kaplan PW, Genoud D, Ho TW, Jallon P. Etiology, neurologic correlations, and prognosis in alpha coma. Clin Neurophysiol 1999;110:205±213. Kinney HC, Samuels MA. Neuropathology of the persistent vegetative state. A review. J Neuropathol Exp Neurol 1994;53(6):548±558. Lemmi H, Hubbert CH, Faris AA. The electroencephalogram after resuscitation of cardiocirculatory arrest. J Neurol Neurosurg Psychiatry 1973;36:997±1002. Lindgren S, Petersen I, Zwetnov N. Prediction of death in serious brain damage. Acta Chir Scand 1968;134:405±416. MoÈller M. `Alpha-Pattern Coma' and survival after cardiac arrest. Electroenceph clin Neurophysiol 1978;44:518±522. MoÈller M, Holm B, Sindrup E, Nielsen BL. Electroencephalographic prediction of anoxic brain damage after resuscitation from cardiac arrest in patients with acute myocardial infarction. Acta Med Scand 1978;203:31±37. Mullie A, Buylaert W, Michem N, et al. Predictive value of Glasgow Coma score for awakening after out-of-hospital cardiac arrest. Lancet 1988;1:137±140. Obeso J, Iragui MI, Marti-Masso J, et al. Neurophysiological assessment of alpha pattern coma. J Neurol Neurosurg Psychiatry 1980;43:63±67. Sacco RL, Van Gool R, Mohr JP, et al. Non traumatic coma. Glasgow Coma score and coma etiology as predictors of 2-week outcome. Arch Neurol 1990;47:1181±1184.

Scollo-Lavizzari G, Bassetti C. Prognostic value of the EEG in post-anoxic coma after cardiac arrest. Eur Neurol 1987;26:161±170. Scott DF, Sumra RS. Unusual EEG patterns in coma, and their evolution. J Neurol Neurosurg Psychiatry 1985;48:80±83. Snyder BD, Ramirez-Lassepas M, Lippert DM. Neurologic status and prognosis after cardiopulmonary arrest: I. A retrospective study. Neurology 1977;27:807±811. Snyder BD, Gumnit RJ, Leppik IE. Neurologic prognosis after cardiopulmonary arrest: IV. Brain-stem re¯exes. Neurology 1981;31:1092±1097. Soerensen K, Thomassen A, Wernberg M. Prognostic signi®cance of alpha frequency EEG rhythm in coma after cardiac arrest. J Neurol Neurosurg Psychiatry 1978;41:840±842. Synek VM. Value of a revised EEG coma scale for prognosis after cerebral anotia and diffuse head injury. Clin Electroenceph 1990;21:25±30. Synek VM, Synek BJL. Theta pattern coma, a variant of alpha pattern coma. Clin Electroenceph 1984;15:116±121. Synek VM, Synek BJL. `Theta pattern coma' occuring in younger adults. Clin Electroenceph 1987;18:54±60. Synek VM, Synek BJL. Transition from alpha to theta pattern coma in fatal cerebral anoxia. Clin Exp Neurol 1988;25:109±113. Trojaborg W, JoÈrgensen EO. Evoked cortical potentials in patients with `isoelectric' EEGs. Electroenceph clin Neurophysiol 1973;35:301±309. Uldry PA, Despland PA, Regli F. Alpha-coma: preÂsentation reÂtrospective de 20 cas. Neurophysiol Clin 1991;21:85±94. Vignaendra V, Wilkus RJ, Copass MK, Chatrian GE. Electroencephalographic rhythms of alpha frequency in comatose patients after cardiopulmonary arrest. Neurology 1974;24:582±588. Walser H, Mattle H, Keller HM, Janzer R. Early cortical median nerve somatosensory evoked potentials. Prognostic value in anoxic coma. Arch Neurol 1986;42:32±38. Westmoreland BF, Klass DW, Sharbrough FW, Reagan T. Alpha-coma. Arch Neurol 1975;32:713±718. Young GB, Blume WT, Campbell VM, et al. Alpha, theta, and alpha-theta coma: A clinical outcome study utilizing serial recordings. Electroenceph clin Neurophysiol 1994;91:93±99. Zandbergen EGJ, de Haan R, Stoutenbeek CP, et al. Systematic review of early prediction of poor outcome in anoxic±ischaemic coma. Lancet 1998;352:1808±1812. Zeman A. Persistent vegetative state. Lancet 1997;350:795±799.