SEARCH

SEARCH BY CITATION

Keywords:

  • Postictal EEG suppression;
  • Sudden unexpected death in epilepsy;
  • Tonic–clonic seizures;
  • Epilepsy

Summary

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References

Sudden unexpected death in epilepsy (SUDEP) has on rare occasions occurred during electroencephalography (EEG) telemetry, and in such cases postictal EEG suppression (PI EEG-SUP) was frequently observed. More recently a retrospective case–control study reported this pattern as a risk factor for SUDEP. We retrospectively audited frequency and electroclinical features of this pattern as well as immediate management following tonic–clonic seizures during telemetry. Forty-eight patients with tonic–clonic seizures were identified from 470 consecutive EEG-videotelemetry reports. Thirteen patients (27%) with PI EEG-SUP (mean duration 38.1 s, range 6–69 s, median 38 s) were compared to 12 randomly selected controls. One seizure was analyzed per individual. Those with PI EEG-SUP were significantly more likely to be motionless after the seizure and have simple nursing interventions performed (suction, oxygen administration, placed in recovery position, vital signs checked). This pattern is relatively common and requires further study as a potential marker for increased mortality in epilepsy.

Patients with epilepsy carry an increased risk of sudden death compared to the general population without epilepsy. This is thought to be related to periictal respiratory and cardiac compromise. Incidence rates range from 0.35 of 1,000 in a population-based study (Ficker et al., 1998) to about 1:100 in the most intractable cohorts (Tomson et al., 2008). Case–control studies in sudden unexpected death in epilepsy (SUDEP) identified a number of associated risk factors including uncontrolled generalized tonic–clonic seizures (GTCS), frequent seizures, never having been treated, polytherapy and frequent medication changes, number of previous antiepileptic drugs (AEDs) tried, learning disability, and lack of supervision (Tomson et al., 2008). Despite this, assessment of individual susceptibility remains difficult. In relatively rare recorded SUDEP and near SUDEP cases (Dasheiff, 1991; Thomas et al., 1996; Bird et al., 1997; Lee et al., 1999; So et al., 2000; McLean & Wimalaratna, 2007; Tavee & Morris, 2008; Bateman et al., 2010) postictal generalized EEG suppression has been observed in a proportion. Although well recognized, there has been only one systematic study of this pattern. Lhatoo et al. (2010) carried out a retrospective study of 10 adult patients (30 seizures) who had undergone EEG video recording and who later died of SUDEP comparing them to a control group. They reported postictal EEG suppression in 15 of 30 case seizures (50%) and 35 of 92 control (38%) seizures. They concluded that prolonged postictal electroencephalography (EEG) suppression (>50 s) appears to identify patients with refractory epilepsy who are at risk of SUDEP and that this risk may be increased in direct proportion to the duration of postictal suppression.

We audited the records of patients with epilepsy who had undergone EEG video-telemetry to (1) describe the frequency of postictal EEG suppression (PI EEG-SUP), in primary or secondary GTCS; (2) to compare associated electroclinical features in a random group of patients without PI EEG-SUP; and (3) to review immediate management.

Methods

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References

PI EEG-SUP was defined as unilateral or bilateral EEG suppression after seizure termination with an amplitude <10 μV. Consecutive telemetry reports over a period of 15 months between October 2007 and December 2008 were reviewed, and all reports mentioning recorded GTCS were identified. Digital video-EEG recordings were reviewed by one of the authors (MS) with regard to immediate postictal EEG patterns following GTCS, and cases with PI EEG-SUP were identified. Electroclinical features were ascertained in those with PI EEG-SUP and a randomly selected group with GTCS without PI EEG-SUP. One seizure with a clear EEG recording and video was analyzed per person. Response of staff was noted by careful video review.

Results

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References

Of 470 patients undergoing EEG videotelemetry during the period of the audit, 48 (10.2%) had 85 GTCS. In 13 (27%) of the 48 cases, immediate PI EEG-SUP was observed, and was generally seen in other seizures if more than one GTCS was recorded per individual. Characteristics of the epilepsy and clinical features in those with PI EEG-SUP and 12 randomly selected cases with GTCS without this pattern, all of whom had postictal slowing (PI EEG-SLOW), are listed in Table 1. Although there were more male than female patients in the PI EEG-SUP group, there was no significant difference between the two groups in sex, age, epilepsy syndrome, drug reduction, polytherapy, or number of seizures recorded.

Table 1.   Patient characteristics
 All Mean, median, range No (maximum total = 25), %PI EEG-SLOW Mean, median, range No (maximum 12), %PI EEG-SUP Mean, median, range No (maximum 13), %
Age30, 32 (8–53)32, 35 (8–51)29, 26 (10–53)
SexF (44%) M (56%)F 50% M 50%F 39% M 69%
Years of epilepsy15, 12 (2–41) (n = 23)14.7 (2–41) (n = 11)15.15 (3–33) (n = 12)
Epilepsy syndrome   
 Focal epilepsy19/25 (76%)9/12 (75%)10/13 (77%)
  Temporal lobe16/19 (84%)7/9 (78%)9/10 (90%)
 Generalized epilepsy4/25 (16%)2/12 (17%)2/13 (15%)
 Not classified2/251/121/13
Other   
 Polytherapy13/22 (59%)7/10 (70%)6/12 (50%)
 Drug reduction20/21 (95%)11/11 (100%)9/10 (90%)

One seizure was analyzed per individual selected on the basis of having a good quality EEG recording and video (Table 2). PI EEG-SUP was observed to start suddenly and bilaterally in all cases and was followed after a variable period by the gradual emergence of slow waves. In 5 of 13 patients, PI EEG-SUP was first noted near the end of the ictal phase associated with a period of stillness between clonic jerks electrographically synchronous with bursts of polyspikes. The onset of EEG suppression at seizure termination was associated with the comatose patient becoming motionless in 12 of 13 patients, one of whom then had brief slow flexion in one upper limb. In the remaining patient the motionless period preceded EEG suppression by 1 min 30 s. In 9 of 12 patients with PI EEG-SLOW, spontaneous movement was noted on video in the immediate postictal period. The duration of PI EEG-SUP varied from 6–69 s (mean 38.1 s, median 38 s). Among the 25 seizures in both groups, only two of the PI EEG-SLOW group showed a prompt early reaction to verbal stimulation, obeying commands, or answering questions. Responses to the other stimulation including vocalization, eye opening, or movement were noted (Table 2). Although respiratory measurements were not undertaken, it was possible to ascertain that most patients in both groups were breathing. One patient in the PI EEG-SUP group appeared to be apneic for about 10 s, and breathing could not be ascertained in another. However, although clearly making respiratory effort, 9 of 11 of the remaining patients in the PI EEG-SUP group and 9 of 12 of the PI EEG-SLOW groups were noted to have altered, potentially compromised respiration, as indicated by airway noise and/or variable respiratory pattern. Specific responses by the nursing staff seen on video, including suction, administering oxygen, placing the patient in the recovery position, or checking vital signs (Table 2) were observed significantly more often in the PI EEG-SUP group compared to the PI EEG-SLOW group (p = 0.012). No serious arrhythmia was documented in either group. Three of the 13 PI EEG-SUP patients in the group are known to have died since the recording. One patients died suddenly (reportedly unresponsive after a 20-min witnessed seizure with no cause found at autopsy), the second died of a head injury secondary to a seizure, and the third of an unrelated illness. Seizure durations and length of PI-EEG-SUP are shown in Table 3.

Table 2.   Electroclinical features in PI EEG-SUP and controls
 All No (maximum = 25),%PI EEG-SLOW No (maximum = 12), %PI EEG-SUP No (maximum = 13), %
  1. Fisher’s exact test. Values in bold indicate statistical significance.

  2. *p = 0.001, **p = 0.012.

No of GTCS: seizures/individual56 in 25 individuals28 in 12 individuals28 in 13 individuals
Mean, median (range)2.2, 2, (1–9)2.3, 2, (1–9)2.15, 1, (1–5)
Electroclinical features of representative seizure per individual   
 Spontaneous movement at seizure termination10/25 (40%)9/12 (75%)*1/13 (8%)*
Response to stimulus if applied (see text)   
 Verbal2/22 (9%)2/10 (1 by parent)0/12 (1 by parent)
  Obeyed command220
 Light touch3/122/41/8
 Moving person0/60/10/5
 Suction1/9 (11%)0/31/6 (17%)
 Individual response to any stimulus4/22 (18%)2/10 (20%)2/12 (17%)
Abnormal breathing (see text)21/25 (84%)9/12 (75%)12/13 (92%)
Measures taken by staff where attended20/259/1211/13
 Suction9/20 (45%)3/9 (33%)6/11 (55%)
 Oxygen6/20 (30%)1/9 (11%)5/11 (45%)
 Recovery position6/20 (30%)1/9 (11%)5/11 (45%)
 Vital signs2/20 (10%)0/9 (0%)2/11 (18%)
 Minimum one measure per case14/20 (70%)5/9 (56%)9/11 (82%)
 Proportion where measures applied23/805/36**18/44**
Table 3.   Duration of analyzed seizure and that of postictal EEG suppression in minutes:seconds
  1. Correlation coefficient: All data: −0.28; ignoring outlier (*) +0.42.

Duration of seizureDuration of PI EEG-SUP
01:0800:28
01:1500:17
01:0700:23
07:0700:06*
01:1801:04
00:5901:05
02:0001:09
02:5400:54
02:1600:38
01:1900:47
00:4800:11
01:3100:54
01:0900:20

Discussion

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References

SUDEP is usually unwitnessed but there is evidence that it is most often due to periictal cardiorespiratory compromise related to GTCS (Bateman et al., 2008; Tomson et al., 2008; Seyal & Bateman, 2009). Respiratory changes are important, and supervision outside of a hospital setting may provide some protection (Langan et al., 2005). A number of risk factors for SUDEP have been identified but more work is needed with respect to individual susceptibility. In the relatively rare SUDEP cases occurring during EEG monitoring, postictal generalized EEG suppression has been observed to be a feature (Tomson et al., 2008). Furthermore, the retrospective case–control study by Lhatoo et al. (2010) provides evidence that prolonged postictal EEG suppression is a marker for increased risk of SUDEP.

Early PI EEG-SUP was observed relatively frequently after GTCS, which occurred in 13 (27%) of 48 individuals. PI EEG-SUP gradually gave way to postictal slowing. We are unable to comment on the late occurrence of this pattern, as saved video-EEG clips were generally only a few minutes in duration. PI EEG-SUP was bilateral and significantly associated with being motionless, occurring in 12 of 13 cases with EEG suppression compared to 4 of 12 with postictal slowing. The cause of postictal EEG suppression is uncertain but could be related to endogenous mechanisms involved in seizure termination (Lado & Moshé, 2008).

Of particular interest is the observation that, compared to a control group (PI EEG-SLOW), cases with PI EEG SUP were significantly more likely to have interventions such as suction, oxygen administration, being placed in the recovery position, or have vital signs checked by nursing staff, suggesting that their clinical postictal state gave rise to more concern in the absence of any serious arrhythmias. The nurses are not EEG trained. The three deaths we are aware of all had PI EEG-SUP. However, without knowing the status of all other cases in the PI EEG-SUP, we are unable to compare mortality between the two groups.

This audit has limitations. Numbers are small, assessment was not blinded, and no data were collected regarding seizures that preceded the index seizure. Nevertheless, we believe that the observations in this retrospective audit need to be followed up with a prospective case–control cohort study that investigates cardiorespiratory parameters and responsiveness during PI EEG-SUP, the effect of physical stimulus on postictal state and, importantly, whether this pattern identifies patients at higher risk of SUDEP.

Disclosure

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References

None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

References

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References