Abstracts

Rationale: Single pulse electrical stimulation (SPES) has been used by several investigators to evaluate effective connectivity (EC) in functional networks, such as the language, visual and limbic systems^1,2,3. The purpose of this study was to apply directional information from EC to gain better understanding of how regions within the seizure onset zone (SOZ) may have widespread network effects. Methods: : SPES was performed in SOZ and outside the SOZ (non-SOZ).  SOZ was defined  as  the  first  EEG  pattern  consisting  of  rhythmic  spikes,  rhythmic  sharp  waves,  recruiting  gamma  activity,  regular  or  low-amplitude  activity  in  the  beta  range  prior  to  or  coinciding  with  the  seizure clinical  manifestation⁴. Two board certified epileptologists (NEC,JYK) reviewed EEG to verify SOZ. SPES was conducted in bipolar  fashion through  a  pair  of  adjacent  electrodes with following stimulation  parameters: monophasic,  alternating  polarity,  square  wave  pulse  with  pulse  width  of  0.3  ms  at  fixed  frequency  of  0.2  Hz, average  of  50  trials /electrode pair.  Current  intensity  was titrated  to point where local/distant  evoked  response (ER)  seen during  real-time  analysis.  If  local/distant  ER  was not  seen  during  real-time  analysis, current  intensity was titrated up  to  80%  maximum  intensity  that  had  a  response at 50  Hz  cortical  stimulation  or  to 12  mA.  For  each  response  site,  ER  amplitude  was calculated  based  on  the  peak  amplitude  response  relative  to  the  pre-stimulus  baseline  (analysis  window:  -500  to  1500  ms).  The  first  10  ms  following  stimulation  was  excluded  from  analysis  to  discard change  due  to  stimulation  artifact.  We calculated the z score (Z= (X - µ) / s) of the amplitude of the ER in reference to pre-stimulus baseline (-500 to -10 ms). Any recording site with z > 6 lasting for at least 30 ms was considered a significant region of connectivity. To assess for degree of influence, for each electrode we assessed (1) outdegree: total number of significant ERPs (Z>6) elicited by SPES of the electrode, (2) indegree: total number of significant ERPs (Z>6) elicited at the specific electrode by stimulation of other sites. Fisher’s exact (FE) test was used to calculate the difference between two proportions in ER relative to electrode location (SOZ vs. non-SOZ). Mann-Whitney (MW) calculation was used to compare the median number of ER within SOZ vs non-SOZ.  Results: SPES was conducted in 5 implantation cases in 40 SOZ and 58 non-SOZ regions. ER were seen in 59.6 % (35%-81%) of electrodes. In all 5 cases, ER were more likely to be seen within the SOZ vs non-SOZ and statistically significant in 4 cases (RR 1.2-2.7, FE p<0.05). Medium number of ER within the SOZ was significantly higher compared to non-SOZ (MW p<0.0001). There was no statistical difference in number of ER between stimulating SOZ vs non-sOZ. Conclusions: Our results strongly suggest that regions within the SOZ overall have a higher degree of EC, and regions involved in the SOZ demonstrate strong in-degree connections compared to regions outside the SOZ regardless of location (frontal vs mesial temporal).  There was no difference in out-degree in the SOZ vs non-SOZ. Our results provides important insight into organization and directedness of seizure networks; the SOZ may be highly connected sub-network that may serve as integrators for network hyper-excitability.   Funding: There was no source of funding that was received in support of this abstract.