Abstracts

Rationale: While intracranial EEG (icEEG) measurements are frequently used for clinical localization of seizure onset, little is known of the relationship between icEEG and other measures of brain function. This is important to better understand the factors which govern intracranial neurophysiology. In this abstract we report on the relationship between icEEG background power, localized MRSI and intracranial microdialysis. This study was performed in patients with hippocampal epilepsy (n=9) in comparison to patients with non-hippocampal (neocortical) epilepsy (n=11). All imaging and intracranial studies were performed in the hippocampus.Methods: All patients were recruited while undergoing seizure localization studies. All hippocampal (Spencer) probes were placed stereotaxically. To minimize post-operative effects, seizure related changes, and to match imaging and intracranial studies, intracranial data were acquired at least 2days after implantation prior medication withdrawal, and at least 6hrs away from any clinical seizures. 1hr long icEEG were acquired with a commercial 128 channel video EEG system. Background power (0.1-50Hz) was measured using an FFT approach. CMA microdialysis probes were used to measure basal inter-ictal levels of GABA (extrapolation to zero-flow); levels of GABA were analyzed with HPLC. MRS studies were performed up to 2mo prior to intracranial study at high field (4T, volume TEM coil transceiver) with moderate echo (TE/TR 72msec/2sec) spectroscopic imaging studies (0.64cc, 19min). Because of non-normal distribution of EC GABA, the correlations with metabolic data were analyzed after log transformation of the microdialysis data. Pearson correlation analysis was used with p<0.02 for statistical significance. Not all patients who underwent icEEG had microdialysis performed.Results: Analysis of background power with hippocampal NAA/Cr demonstrated a significant exponential relationship in the MTLE group (R=+0.90, p<1xe-4, Fig. 1); there was no significant relationship in the non-MTLE group. Log ecGABA significantly correlated with background power in the MTLE group (Fig. 2, R=+0.93, p<0.003) and in the nonMTLE group (R=-0.88, p<0.01). Conclusions: In MTLE, the background EEG power is exponentially correlated with the MRSI measure of mitochondrial function. That such a relationship is not seen in the nonMTLE group suggests that under more normal conditions, mitochondrial function is not limiting for basal electrical activity. In MTLE, the tight coupling between background power and mitochondrial function may be via GABA; e.g., it is known that mitochondrial injury results in increased GABA release. However Fig. 2 shows that in MTLE, increased GABA is associated with increased EEG power. This may indicate either a failure of GABA to suppress electrical activity, or as has been suggested, be an excitatory effect of GABA. Sources of funding: NIH EB00473, NS54038