NEONATAL SOCIETY ABSTRACTS
The relationship between best clinical estimate of the time of cerebral injury in the newborn and EEG seizure onset
Presented at the Neonatal Society 2003 Summer Meeting (programme).
Filan P, Chorley G, Boylan G, Davies A, Pressler R, Fox G, Rennie J
4th Floor, Golden Jubilee Wing, King's College Hospital, LONDON SE5 9RS
Background: In a lamb model of brain injury an acute insult was followed by a period of EEG depression which started after 10-20 minutes and lasted for approximately 8 hours. After this seizures emerged, at around 10±3 hours (1). Even in the animal model there is wide variation in the duration of insult which is required to produce an isoelectric EEG (2), and human babies are not subjected to identical and reproducible insults. Despite this, we felt that it would be interesting to examine the relationship between the EEG determination of seizure onset time and our best estimate of the time of any fetal/ neonatal insult in the human baby.
Aims: To establish whether or not the time of seizure onset, as determined by EEG monitoring, was related to the time of the fetal/neonatal cerebral insult (3). Previous work has examined the time of clinical seizure onset, which is unreliable when compared to EEG.
Methods: Babies who were considered to be at high risk of seizure were monitored prospectively using 12 channel continuous video-EEG telemetry, with parental consent. The best estimate of the time of brain injury was determined by careful review of the CTG traces, the obstetric history, and the peripartum events by an observer (AD) who was blind to the EEG findings. In one case there were neuropathological findings from autopsy.
Results: Eight babies with hypoxic-ischaemic encephalopathy had monitoring commenced sufficiently early to allow detection of the time of first electrographic seizure, and a further baby was already seizing when monitoring began (at 5 hours). In 3 babies it was thought that the insult had occurred before labour and more than 12 hours before birth. In this group the EEG seizure onset time was 4, 5.5 and 9.5 hours respectively. These babies began seizing earlier than the 6 whose insult was thought to have occurred during labour or close to the time of birth. In these babies the first seizure was detected at 11.5, 12, 13, 18, 20 and 26 hours after birth. This gave a median seizure onset time of 15.5 hours compared to 5.5 hours (Mann Whitney P value=0.02).
Conclusion: Babies who had suffered an acute hypoxic-ischaemic insult close to birth had their first EEG seizure later than those whose history suggested that their brain injury was sustained before labour began. The EEG in this group followed a very similar pattern to that seen in the lamb model after an acute insult. Our results suggest that babies with EEG seizure onset at less than 6 hours probably did not sustain their brain injury at around the time of birth. Continuous EEG monitoring from birth may allow a more precise estimate of the time of brain injury to be made than that which is currently possible with other methods.
1. Williams CE, Gunn AJ, Mallard C, Gluckman PD 1992 Outcome after ischemia in the developing sheep brain: an electrographic and histological study. Annals of Neurology 31:14-21.
2. Mann LI, Prichard, JW, Symmes D 1970 EEG, ECG, and acid-base observations during acute fetal hypoxia. American Journal of Obstetrics & Gynecology. 106:39-51
3. Hollier LM 2000 Can neurologic injury be timed? Seminars in Perinatology 24:204-214