Epilepsia, 47(2):343–354, 2006 Blackwell Publishing, Inc. C  2006 International League Against Epilepsy Temporal and Extratemporal BOLD Responses to Temporal Lobe Interictal Spikes Eliane Kobayashi, Andrew P. Bagshaw, Christian-George B´ enar, Yahya Aghakhani, Frederick Andermann, Fran¸ cois Dubeau, and Jean Gotman Montreal Neurological Institute and Hospital, McGill University, Montreal, Quebec, Canada Summary: Purpose: Simultaneous EEG and functional MRI (f MRI) allows measuring metabolic changes related to interictal spikes. Our objective was to investigate blood oxygenation level– dependent (BOLD) responses to temporal lobe (TL) spikes by using EEG-f MRI recording. Methods: We studied 35 patients who had a diagnosis of tem- poral lobe epilepsy (TLE) and active TL spiking on routine scalp EEG recording. Two-hour sessions of continuous EEG-f MRI were recorded, and spikes were identified after offline artifact removal and used as events in the f MRI analysis. Each type of spike was analyzed separately, as one EEG-f MRI study. We de- termined significant (p < 0.05) positive (activation) and negative (deactivation) BOLD responses for each study. Results: Twenty-seven patients had spikes during scanning (19 unilateral and eight bilateral). From a total of 35 f MRI studies, 29 (83%) showed BOLD responses: 14 had both activations and de- activations; 12, activations only; and three, deactivations only. Six (17%) showed no responses. Nineteen studies had mainly neocortical TL activation: Sixteen (84%) of 19 concordant with spikes, 12 of 16 with concomitant activation of the contralat- eral TL, and 16 of 19 with additional extratemporal activation; few showed exclusively mesial TL activation. Seventeen studies showed deactivation, either extratemporal plus temporal (n = 8) or exclusively extratemporal (n = 9). Conclusions: BOLD responses to TL spikes occurred in 83% of studies, predominated in the spiking temporal lobe, and manifested as activation or deactivation. Responses often in- volved the contralateral homologous cortex at the time of uni- lateral spikes and were frequently observed in extratemporal regions, suggesting that TL epileptic spikes can affect neu- ronal activity at a distance through synaptic connections. Key Words: EEG-f MRI—TLE—Interictal temporal lobe spikes— BOLD response. Temporal lobe epilepsy (TLE) is the prototype of localization-related epilepsy. Nevertheless, evidence for more diffuse changes in this disorder, involving cortical and subcortical structures outside the epileptogenic tem- poral lobe (TL), is provided from a multiplicity of sources. Surgical results in TLE are good but far from perfect, particularly in long-term follow-up studies (1,2). Electro- physiological studies point to regions outside the epilep- togenic TL in some individuals, particularly the insula (3) and perisylvian cortex (4). Volumetric magnetic resonance (MR) measurements suggest involvement of contralateral TL, thalami, and striatum (5,6). Positron emission tomog- raphy (7–12) and MR spectroscopy (13,14) also point to widespread abnormalities. These functional neuroimag- ing studies can provide a broad view of the metabolic dysfunction throughout the brain, but they do not have the Accepted August 23, 2005. Address correspondence and reprint requests to Dr. E. Kobayashi at Department of Neurology and Neurosurgery, Montreal Neurological In- stitute and Hospital, McGill University, 3801 University Street, Montreal Que., Canada, H3A 2B4. E-mail: eliane.kobayashi@mail.mcgill.ca temporal resolution to assess metabolic changes directly related to epileptic spikes. Combined EEG and functional MR imaging (EEG- f MRI) can identify positive and negative changes in the blood oxygenation level–dependent (BOLD) f MRI con- trast related to focal and generalized interictal epileptic spikes. The significance of BOLD responses in terms of specific neuronal activity is still under investigation (15,16). This technique is able, however, to provide a broad view of the metabolic changes in the brain with high temporal resolution (on the order of a second), and hence provides a “snapshot” of the functional state of the entire brain at the time of an epileptic discharge. Previous studies exploring EEG-f MRI findings in pa- tients with partial epilepsies (17–24) demonstrated that f MRI responses could be obtained in regions that corre- late with localization of the interictal EEG spikes, but also in other distant brain regions. These responses were seen in the subcortical regions and in the areas homologous to the spikes. This interesting finding suggests that focal epilepsies may not be as focal as previously assumed (25– 27). Also, most studies reported positive BOLD responses 343