Relation of interictal spike frequency to H-1-MRSI-measured NAA/Cr

Purpose: Whereas EEG spiking and decreases of the neuronal marker N-acetyl- aspartate (NAA) both localize well the epileptic focus, the significance of the intensity of these variables is unclear. Therefore we investigated whe ther the frequency of interictal surface spikes is related to the degree of N-acetyl-aspartate/creatine (NAA/Cr) ratio decrease as measured by proton magnetic resonance (MR) spectroscopic imaging (H-1-MRSI) in patients with i ntractable partial epilepsy. Methods: We retrospectively studied 14 patients, nine with temporal lobe ep ilepsy and five with frontal lobe epilepsy. Spikes that occurred during pro longed video-EEG monitoring from electrodes placed according to the Interna tional 10-20 system were counted blinded to the H-1-MRSI results. Eight ele ctrode positions (F3/4, C3/4, T3/4, T5/6) were assigned to underlying brain subregions in the H-1-MRSI volume of interest. We converted NAA/Cr ratios into z-scores (NAA/Cr,) to compared NAA/Cr values directly across subregion s. We calculated Spearman rank-order (p) and Pearson product-moment (r) cor relations between spike frequency and NAA/Cr, values overall, as well as wi thin each brain subregion. Results: We found an overall negative relation between spike-frequency data and NAA/Cr, data (p = -0.341). When analyzing only spiking subregions, thi s negative relation be came slightly stronger(p = -0.442; r = -0.338). When data from the eight sites were considered separately, this negative relati on remained in most instances. Conclusions: Our results reveal a trend toward higher interictal spike freq uencies on surface EEG in regions of pronounced neuronal metabolic damage o r dysfunction. This suggests that both variables parallel an underlying pat hologic substrate, although the pathophysiologic processes may be distinct.