Abstracts

This is a Late Breaking abstract

Rationale: Previous research has reported inconsistent results on the relationship between inter-ictal spiking in electroencephalography (EEG) and epileptogenesis after brain injury. We tested the hypothesis that interictal spike frequency changes over time after brain injury, regardless of whether a subject develops epilepsy.

Methods: Rats received bilateral epidural wire electrode implants and were monitored with a two-channel wireless EEG system. Using the Rice-Vannucci model at post-natal day 30, 43 rats were subjected to either hypoxia-ischemia (HI) for 2 hr or sham-control brain injury. In most subjects, this procedure did not consistently result in stroke; however, many animals in the study were found to have various brain abnormalities and injuries. Spontaneous recurrent seizures (i.e., epilepsy) were not detected in any of the rats, but some exhibited early electrographic seizure(s). For inclusion, 25 rats were selected based on a minimum duration of 75 days of longitudinal electrical recordings. Randomly selected and merged EEG segments from representative days 5, 40, and 75 after HI were extracted and anonymized for each rat. Discharges defined as “simple inter-ictal spikes” with duration 20-70 msec were narrowed down using Python and C++ script(s) and verified via manual human interpretation. Animal MRIs were also anonymized for interpretation and injury classification.

Results: A total of 1197 discharges were included in the analysis. We found that the frequency of inter-ictal spike onsets generally increased over time after brain injury (p=0.007 via 2-tailed Wald Test with the null hypothesis of the linear regression slope “m” = 0). The trajectories or “slopes” of spike frequencies were clustered via k-means++ (k=3) to stratify the patterns of injury type, and were also analyzed in conjunction with MRI findings. MRIs showed various injury types, including penetrating electrode injuries, brain tumor(s), and abscesses. A severity score was developed using T1 and T2 weighted MRIs. Multiple linear regression coefficients for the relationship between severity scores of different injury types and EEG spike frequency slopes were computed in 13th dimensional space (as there were 13 injury types, with subsets of them co-occurring simultaneously). Our exploratory analysis has not yet found any relationship between the total severity score for the injury types and the change in spike frequency over time.

Conclusions: Our results suggest that several types of brain injury are followed by inter-ictal spiking, even in the absence of evidence for epilepsy. Spike frequency steadily increased over the 75-day interval in a subset of rats. Although further exploratory analysis has not found any obvious relationship between the severity or type of brain injury and change in spike frequency, further examination of MRI data is pending. Thus, our results suggest that interictal spiking alone is not a biomarker for epileptogenesis, and that interictal spiking may be a common result for brain injury in general.

Funding: Supported by NINDS 5R01NS086364 and DOD-CURE contract W81XWH-15-2-0069
Conflicts of interest: F.E. Dudek has equity interest in and received remuneration from Epitel, Inc. and Cage Data Corp for work related to wireless EEG devices for rodents and humans. A. Zayachkivsky received consulting fees from Cage Data Corp. and BIOPAC Systems for work on the EPOCH telemetry device.