FULL-LENGTH ORIGINAL RESEARCH
Abnormal response to photic stimulation in Juvenile Myoclonic Epilepsy: An EEG-fMRI study *Emanuele Bartolini, †Ilaria Pesaresi, *Serena Fabbri, *Paolo Cecchi, ‡Filippo Sean Giorgi, *Ferdinando Sartucci, *Ubaldo Bonuccelli, and §Mirco Cosottini Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
SUMMARY
Emanuele Bartolini is a Research Fellow at the University of Pisa, Italy.
Objective: Juvenile myoclonic epilepsy (JME) is a young-onset electroclinical syndrome, characterized by myoclonic, generalized tonic–clonic, and possibly typical absence seizures. Interictal electroencephalography (EEG) displays 3–6 Hz spike/ polyspike and wave pattern. Photosensitivity is common. Our aim was to explore the blood oxygen level–dependent (BOLD) response evoked by a highly provocative photic stimulus in a cohort of people with JME compared to a group of nonphotosensitive healthy controls, and to investigate the hemodynamic phenomena seen in patients with photosensitive JME. Methods: We studied 13 JME patients and 18 healthy controls using EEG–functional magnetic resonance imaging (fMRI) performed during low luminance intermittent photic stimulation (IPS). The BOLD response to IPS was investigated both in JME and control groups. In photosensitive JME subjects, we also performed a dynamic evaluation of BOLD signal changes evoked by the photoparoxysmal response (PPR) in a time frame ranging from 10 s before the onset of the EEG paroxysm up until 10 s afterward. Results: The IPS evoked a positive BOLD response in striate and extrastriate visual areas, which was less in JME patients than in controls. Moreover, people with JME had a reduced positive BOLD response in the frontoparietal areas and putamen but a stronger negative BOLD response in the primary sensorimotor cortex (SM1) and in cortical regions belonging to the default mode network (DMN). In JME, the dynamic evaluation of BOLD signal changes related to PPR revealed an early positive response in the putamen and SM1, followed by BOLD signal decrements in the putamen, caudate nuclei, thalami, and SM1. Significance: Our results confirm the hypothesis that people with JME might have an altered interaction between the motor circuit and other neuronal networks, with prominent involvement of basal ganglia circuitry. The PPR could be a final expression of pathogenic phenomena occurring in the striato-thalamocortical system, possibly a core feature of system epilepsy JME. KEY WORDS: Juvenile Myoclonic Epilepsy, Neuroimaging, Neurophysiology, Photosensitivity.
Accepted March 19, 2014; Early View publication May 23, 2014. *Department of Clinical and Experimental Medicine, University of Pisa, Italy; †Unit of Neuroradiology, Santa Chiara Hospital AOUP, Italy; ‡Unit of Neurology, Santa Chiara Hospital AOUP, Pisa, Italy; and §Department of Translational Research and New Technologies in Medicine and Surgery, University of Pisa, Pisa, Italy Address correspondence to Mirco Cosottini, Department of Translational Research and New Technologies in Medicine and Surgery, University of Pisa, Via Risorgimento 36, 56126 Pisa, Italy. E-mail: mircocosottini @libero.it Wiley Periodicals, Inc. © 2014 International League Against Epilepsy
The recent extensive application of electroencephalography–functional magnetic resonance imaging (EEG-fMRI) in idiopathic generalized epilepsy (IGE) has challenged the common interpretation of syndromes featuring primarily diffuse cortical discharges, and suggests the involvement of complex neuronal networks.1–4 The majority of these articles have addressed the whole IGE group or subgroup constituted by electroclinical phenotypes (e.g., spontaneous generalized spike and waves [GSWs] or absence seizures
1038
1039 A Striato-Thalamocortical Hypothesis [ABS]). No EEG-fMRI study has specifically explored the most diffuse form of IGE, juvenile myoclonic epilepsy (JME). This is a genetically determined syndrome characterized by adolescent/adult-onset myoclonic jerks alone or combined with generalized tonic–clonic seizures (GTCS) and ABS. Interictal EEG is typically characterized by an irregular mixture of 3–6 Hz spike/polyspike and slow wave activity.5 From the neurophysiological point of view, JME is characterized by a lower resting motor threshold than normal and by stronger intracortical c-aminobutyric acid (GABA)ergic inhibition of the motor system in response to photic stimulation.6–8 The latter characteristic is shared with healthy photosensitive subjects.9 Photosensitivity is characterized by abnormalities in the visual system (e.g., altered visual contrast gain control, heightened sensitivity to specific wavelengths).10,11 Whenever a luminous stimulus is sufficiently provocative, a critical mass of occipital neurons are activated, yielding epileptiform activity that may spread toward different areas, especially the frontal lobes.12 JME is deemed to be the most photosensitive form of epilepsy. About one third of patients with JME have a photoparoxysmal response (PPR) during classic stroboscopic intermittent photic stimulation (IPS).13 The ratio could even be higher with more provocative stimulation paradigms, such as low luminance ( 600 nm) flickers.14 We applied EEG-fMRI to study hemodynamic cortical and subcortical response to IPS in a cohort of people with JME compared to a group of healthy subjects. EEG-fMRI is a methodology based on synchronous registration of fMRI and EEG data. It can be used to detect changes in the blood oxygen level–dependent (BOLD) response to external or neurophysiologic events. In this context IPS and PPR occurrence represent phenomena that can be triggered during fMRI, offering a privileged time window to study the associated hemodynamic phenomena. We hypothesize that JME is characterized by a cortical and subcortical network disruption that predisposes to photosensitivity. Starting from this consideration, we investigated the pattern of BOLD response evoked by a highly provocative photic stimulation (deep red low luminance flickers) in people with JME. Furthermore, we aimed to investigate the hemodynamic phenomena that precede and accompany the occurrence of PPR in photosensitive JME patients.
Methods Subjects We enrolled 13 JME patients, selected consecutively from the Epilepsy Clinic of the University of Pisa (mean age 29.3 years, range 16–49 years). Inclusion criteria were the following: (1) definite diagnosis of JME according to the International League Against Epilepsy (ILAE) classification,15 (2) unremarkable findings on conventional brain
MRI scan, and (3) routine EEG with a normal background. Patients’ clinical histories were characterized by myoclonic seizures (n = 13/13), GTCS (n = 12/13), ABS (n = 5/13), and at least one interictal EEG with generalized spike-polyspike and wave (n = 13/13). Demographic and clinical characteristics of the JME cohort are summarized in Table S1. We also recruited 18 healthy controls with no history of febrile convulsions, seizures, or family history for epilepsy. Three healthy controls were excluded for PPR recording during the scan. Included controls were matched with patients for sex (12 female, three male) and age (mean age 26.4 years; range 19–34 years; t-test p-value = 0.2226). The study was approved by the local ethics committee. Written informed consent was obtained from each subject participating in the study, or by the patient’s parents for participants who were younger than 18 years. EEG acquisition, processing, and analysis EEG was continuously recorded within the scanner by using an MRI-compatible cap with 32-electrodes and an MRI-compatible amplifier (BrainAmp-MR; Brain Products, Gilching, Germany). The EEG acquisition and processing is described in detail elsewhere.16 EEG recordings were reviewed offline by an experienced electroencephalographer (E.B.) for identification of EEG abnormalities related to IPS. We analyzed only diffuse PPR, classified as class 4 according to Waltz et al.17 Patients who presented with at least one PPR during the experiment were classified as photosensitive. Photic stimulation inside the scanner In the fMRI experiment IPS was administered in a block design. According to previous work,14 IPS consisted of deep red (wave length 680 nm) low luminance (luminance < 30 cd/m2) flashes of different frequencies (10, 15, and 20 Hz), using 20 s blocks of stimulation alternating with 15 s blocks of rest. IPS blocks of different frequencies were presented in pseudo-random order (every frequency was repeated five times). The subjects were instructed to keep eyes open during EEG-fMRI examination. The light was delivered by an MRI compatible visual stimulation device (VisuaStimResonance Technologies, Northridge, CA, U.S.A.) composed of a fiberoptic goggle system, playing as a liquid-crystal display (LCD) screen placed in front of the eyes. The visual paradigm was managed by the software Presentation (Neurobehavioral Systems, Berkeley, CA, U.S.A.). The EEG was observed online to interrupt the stimulation in case of a seizure or a prolonged PPR. Subjects were instructed to push an emergency button in case of discomfort during IPS. fMRI acquisition fMRI data were acquired continuously during EEG recordings with a 3T scanner (Discovery MR 750; GE Medical System, Milwaukee, WI, U.S.A.). The fMRI data were obtained by T2*-weighted gradient-recalled echo-planarEpilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
1040 E. Bartolini et al. imaging (EPI) sequence (TR 2,500 msec, TE 40 msec, FA 90 degrees, image matrix 128 9 128, in-plane field-of-view 260 9 260 mm2) with 28 interleaved slices (slice thickness 4 mm, gap 1 mm) angled at 30 degrees from the anterior– posterior commissural plane, repeated over 210 volumes for a total scanning time of 8 min 30 s. For each patient, additional high-resolution T1-weighted images were acquired (three dimensional (3D)-BRAVO sequence, TR 8.16 msec, TI 450 msec, TE 3.18 msec, FA 12 degrees, voxel size 1 9 1 9 1 mm, 160 axial-oblique slices, total scanning time 4 min 28 s) to provide accurate anatomic references for functional data.
Both positive and negative BOLD changes were investigated in first-level and high-level statistical analysis.
fMRI data processing Data were analyzed by using fMRI Expert Analysis Tool (FEAT) v5.98 tool of the software-package FSL 4.1.5 (http://www.fmrib.ox.ac.uk/fsl/). Both patient and control datasets underwent the same analysis steps. Each functional dataset underwent preliminary processing including slicescan-time-correction, 3D motion-correction, high-passtemporal-filtering (100 s) and spatial-smoothing (Gaussian Kernel, 8 mm full-width-half-maximum). All subjects showed movement-related displacement lower than 3 mm/ 3 degrees.
Dynamic evaluation of BOLD changes associated with PPR in photosensitive JME patients In photosensitive patients we explored the temporal dynamic of BOLD signal changes associated with PPR by an event-related design. The PPR regressor was modeled with onset and duration of the events and convolved with a gamma-HRF. We explored a time window ranging from 10 s before the onset of the EEG paroxysm up until 10 s afterward, by a set of nine GLMs with PPR regressors shifted by 2.5 s in time (Fig. S1). Each GLM also included an IPS regressor (a unique regressor including all the photic stimulation frequencies) and motion displacement parameters as confounds. Group analysis was performed by a fixed effect model to obtain a mean group map of PPR regressor at each temporal bin. In total nine mean group positive BOLD maps and nine mean group negative BOLD maps were obtained. A statistical Z-threshold of 4.0 and cluster p-threshold < 0.05 were applied to all of them.
Comparison of BOLD response evoked by IPS in JME patients and healthy controls First-level statistical analysis was performed by the general-linear-model (GLM) approach. The signal model in each voxel was derived by block-design paradigm with three distinct predictors for each IPS frequency (10, 15, and 20 Hz). The expected BOLD responses were obtained by convolving the original predictor waveforms with a gamma-form hemodynamic response function (standard HRF peak = 6 s). The design matrix of GLM also included motion displacement parameters estimated by Motion Correction using FMRIB's Linear Image Registration Tool (MCFLIRT) as confound variables. For patients with PPR, a dedicated regressor was utilized and added to GLM. To obtain spatially resolved anatomic references for statistical maps, functional datasets were registered to brain-extracted 3D T1-weighted images by linear transformations with 6-degrees-of-freedom (FLIRT tool).18 Then functional and structural datasets were aligned to Montreal Neurological Institute - 152 (MNI-152) standard space (International Consortium for Brain Mapping-152 [ICBM-152] template) by linear 12-degrees-of-freedom transformations (FLIRT tool).19 The higher-level statistical analysis was performed by the fixed-effect model.20 The Z statistical maps resulting from both intragroup analysis and between-group analysis underwent cluster thresholding (Z-threshold = 2.0, cluster pthreshold < 0.05). For between-group statistical analysis in subcortical nuclei, a small volume correction was applied to correct for multiple comparisons (Z-threshold = 2.0, cluster p-threshold < 0.05).21 Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
Comparison of IPS epochs with PPR and without PPR in photosensitive JME patients In photosensitive patients we compared the BOLD signal changes during IPS epochs associated with PPR to epochs without PPR. By detail, IPS blocks with and without PPR were used to build separate regressors in GLM. Both regressors were convolved with gamma-HRF. High-level intragroup analysis was performed by a fixed effect model (Z-threshold = 2.0, cluster p-threshold < 0.05).
Results During EEG-fMRI examinations, IPS evoked at least one PPR in 7 of 13 patients (total number of PPR recorded: 41; mean event duration 1.27 s, range 0.20–4.32 s; Table S1). None of the patients or controls reported discomfort during the experiment. Comparison of BOLD response evoked by IPS in JME patients and healthy controls In healthy subjects, IPS evoked a positive BOLD response in striate and extrastriate visual areas at all frequencies. A positive BOLD response was also observed in frontoparietal areas and in the putamen (especially elicited by 20 Hz IPS). In addition, negative BOLD changes were induced in the primary sensorimotor area (SM1), frontal, parietal, temporal and cingulate cortices (Fig. 1). The pattern of BOLD decrements in frontoparietal areas, temporal cortices, and cingulate cortices resembled the spatial distribution of the so-called default-mode–network (DMN). Patients showed a significantly lower positive BOLD response compared to controls in striate and extrastriate
1041 A Striato-Thalamocortical Hypothesis
Figure 1. BOLD changes evoked by low luminance intermittent photic stimulation in healthy controls. Positive (red-yellow) and negative (bluegreen) BOLD responses evoked by IPS in healthy controls. Epilepsia ILAE
Figure 2. BOLD changes evoked by low luminance intermittent photic stimulation in patients with JME. Positive (red-yellow) and negative (bluegreen) BOLD responses evoked by IPS in patients with JME. Epilepsia ILAE Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
1042 E. Bartolini et al.
Figure 3. Comparison between BOLD changes evoked by low luminance IPS in healthy controls and patients with JME. On the left side: betweengroup differences in positive (red-yellow) BOLD response evoked by IPS. On the right side: between-group differences in negative (bluegreen) BOLD response evoked by IPS. Epilepsia ILAE
visual areas (Brodmann area BA17, 18, 19), in parietal (BA5, 7, 39) and frontal areas (BA6, 9, 44, 46) and in putamen (Figs. 2 and 3 and Table S2). A stronger negativeBOLD response in JME was observed in SM1 (BA1, 2, 3, 4), premotor cortex (BA6), and in the DMN areas (parietal cortex BA39; temporal cortex BA20, 21, 22; frontal cortex BA8, 9, 44, 46; cingulate cortex (BA23, 24, 30, 32; Figs. 2 and 3 and Table S3). Comparison of IPS epochs with PPR and without PPR in photosensitive JME patients In photosensitive patients, IPS epochs associated with PPR were characterized by a weaker positive BOLD response in the striate and extrastriate cortex compared to IPS epochs without PPR. In addition, IPS with PPR elicited a stronger negative BOLD response in SM1, the premotor cortex, DMN, and the putamen (Fig. 4). Dynamic evaluation of BOLD changes associated with PPR in photosensitive JME patients The dynamic analysis of PPR events disclosed an early positive BOLD response in putamen and SM1 accompanied by a negative BOLD response in DMN (0 and +2.5 s from Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
PPR onset). Then, positive BOLD phenomena disappeared and negative BOLD changes surged in the putamen, the head of the caudate nuclei, and the frontomesial cortex (+5 s from the PPR onset). Further on, negative BOLD changes involved the thalami and SM1 (+7.5 s from PPR onset) and then progressively extended to broad cortical regions (+10 s from PPR onset) (Fig. 5).
Discussion The aim of our study was investigating JME pathophysiology by addressing the hemodynamic effects on cortical and subcortical brain structures induced by highly provocative photic stimuli. To build an MRI-compatible IPS paradigm with the maximal epileptogenic potential, we applied photic stimulation with trains of low luminance deep red flickers, generated outside the magnet room and driven to the patient via fibre optics and goggles. Low luminance IPS was studied extensively after the “Pocket Monster incident,” when about 600 Japanese children had seizure while watching a cartoon on a TV screen. The provocative event was identified as a sequence of deep red flickers (wave length >600 nm).22
1043 A Striato-Thalamocortical Hypothesis
A
B
C
D
Figure 4. Comparison between IPS epochs with PPR and without PPR. (A) Positive (red-yellow) BOLD responses induced by IPS epochs without PPR (upper row) and with PPR (lower row). (B) Negative (blue-green) BOLD responses elicited by IPS epochs without PPR (upper row) and with PPR (lower row). (C, D) Differences between BOLD signal changes evoked by IPS epochs with and without PPR (on the left side positive BOLD changes, red-yellow; on the right side negative BOLD changes, blue-green). Epilepsia ILAE
Subsequently Takahashi et al.14 demonstrated that deep red low luminance stimulation is more provocative than conventional stroboscopic IPS performed by high luminance flashes. Photosensitivity is generally regarded as a cortically driven phenomenon, mainly involving the occipital lobes, with a secondary contribution from the frontal areas.23 The development of a PPR probably occurs when normal physiologic excitation of the visual cortex exceeds a critical level as a result of a failure in inhibitory processes. The latter would be characteristic of photosensitive people, possibly as a genetic trait. Consistent with this hypothesis, Strigaro et al.24 have recently demonstrated that photosensitive IGE is characterized by a flaw in the inhibition of the visual system normally induced by a single luminous flash. Previous work has described complementary abnormalities in the visual system of people with photosensitive epilepsy. Porciatti et al.10 reported a defective gain control mechanism of visual stimuli in photosensitive epilepsy, possibly related to defective inhibition, highlighted by visual evoked potentials (VEPs). Takahashi et al. proposed two different pathophysiologic mechanisms of photosensitivity: a “quantity of light dependent” and a “wavelength dependent” mechanism. The first would require stimuli with an elevated
amount of luminance contrast to elicit a PPR; the second would be based on the high provocative properties of low luminance flashes with wavelengths of around 700 nm (i.e., deep red), and would be more often observed in people with IGE and photosensitive healthy subjects.11 The paradigm of stimulation we applied acts by the “wavelength dependent” mechanism. In detail, the color theory states that human visual system carries information about color through antagonistic channels along the parvocellular and magnocellular pathways. These channels derive from the overlapping inputs of three retinal cones (L-cones for red, M-cones for green, S-cones for blue). It has been speculated that low luminance deep red stimuli activate only retinal red cones (L-cones) with no effect on counteracting antagonist cones (M-cones and S-cones), yielding the maximal provocative effect.12 Our stimulus protocol was reinforced by the use of long-lasting stimulation trains, much longer than in former studies.14,22 Elongation of the stimulus train was established because of the previous observation that IPS duration increases the rate of photosensitivity in people with JME.25 We applied three randomordered frequencies, acknowledged as highly provocative (i.e., 10, 15, and 20 Hz).25 The strong provocative effect of Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
1044 E. Bartolini et al.
Figure 5. Dynamic evaluation of BOLD changes associated with PPR in photosensitive JME patients. Positive (red-yellow) and negative (blue-green) BOLD responses related to PPR in JME patients across a 20 s time window (from 10 s before PPR onset up till 10 s after that). Upper rows: axial view at level of basal ganglia. Lower rows: upper view of the cerebral volume rendering. Epilepsia ILAE
the visual stimulation we applied was in fact highlighted by the occurrence of PPR in 3 of 18 healthy subjects. BOLD response evoked by IPS in JME patients and healthy controls People with JME were characterized by a reduced BOLD response to IPS in the striate and extrastriate visual cortex compared to healthy controls. Up until now no definite data have been available on the visual cortex excitability in JME subjects. Previous transcranial magnetic stimulation (TMS) studies demonstrated a heightened excitability of the primary visual area to photosensitivity. In fact, the presence of PPR in healthy subjects is associated with lower phosphene thresholds and steeper stimuEpilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
lus-response curves in studies performed by TMS stimulation over the primary visual cortex.9 Hyperexcitability would affect both magnocellular and parvocellular systems, as demonstrated by pattern-reverse VEPs.26 Furthermore, photosensitive subjects are characterized by stronger habituation of the visual cortex to trains of stimuli, as shown by a stronger inhibitory effect of TMS on visual perception and by a more pronounced VEP habituation to pattern-reverse blocks.9,26 The reduced IPS-evoked positive BOLD response that we observed in patients’ visual cortex could be explained by IPS repetitive nature. Similar to in healthy photosensitive subjects, we hypothesize people with JME could be characterized by a stronger habituation to a repetitive stimulus.
1045 A Striato-Thalamocortical Hypothesis Along with a reduced activation of striate and extrastriate occipital visual cortex, patients presented with a lower BOLD response than controls in vision-related frontoparietal areas. In addition, photic stimulation provided data on motor circuits, unveiling an altered BOLD response in motorrelated cortical and subcortical areas. In detail, a stronger BOLD signal decrease was observed in the SM1 and premotor cortex of patients compared to controls, with maximum effect at 10 Hz frequency. The interpretation of such results is not straightforward. Studies performed by TMS have previously demonstrated that photic stimulation normally increases the motor cortex excitability, thereby shortening the cortical silent period (CSP) in healthy not-photosensitive subjects. On the contrary, IPS does not modify CSP either in healthy photosensitive subjects or in people with IGE/JME (whether photosensitive or not).7–9 CSP is a measure of inhibitory processes within the motor cortex, mediated by GABA neurons. Resting motor threshold (rMT) is instead believed to reflect membrane excitability of corticospinal neurons and related interneurons within the motor cortex. Siniatchkin et al.9 studied photosensitive and nonphotosensitive healthy subjects and did not detect any rMT differences between the groups. Consequently we may speculate that the differences in rMT can be attributed to factors other than the photosensitive trait. Of interest a recent meta-analysis by Brigo et al.6 has assessed that drug-naive people with JME have a lower rMT and a heightened motor excitability compared to healthy subjects. Hence, the sensorimotor cortex of people with JME would be characterized by a major excitability balanced by stronger GABAergic intracortical inhibition. The stronger BOLD deactivation in SM1 that we detected in patients could express a more intense GABAergic intracortical inhibition. With regard to subcortical structures, IPS induced a positive BOLD response in the putamen nuclei of healthy controls, which was lacking in patients. A role of the putamen in visual-motor integration has formerly been hypothesized, with the description of different classes of neurons responding specifically to pure visual stimuli next to neurons involved in motor planning/inhibition.27 Dysfunction of basal ganglia circuitry has already been reported in epilepsy.28 Specifically JME has been associated with microstructural and macrostructural abnormalities in the putamen29 and with dopaminergic dysfunction. In particular, PET studies have demonstrated a reduced binding to the dopamine transporter in the substantia nigra and midbrain and reduced binding to dopamine D2/D3 receptors in the posterior putamen.30,31 An abnormal BOLD response to IPS was also observed in areas belonging to the DMN, in particular a stronger BOLD signal reduction was detected in JME patients compared to controls. Normally DMN is anticorrelated to the sensorimotor network.32 On the contrary we observed people with JME
had deactivation in both systems. A previous study has already described an alteration of DMN in JME patients, suggesting that DMN impairment could lead to hyperconnectivity across systems, including the motor-related areas.33 Because DMN abnormalities have been reported in many different epileptic phenotypes independently of photosensitivity,2,4,33 it is difficult to establish whether DMN could play a causal role or it represents a not specific trait. Comparison of IPS epochs with PPR and without PPR in photosensitive JME patients In the photosensitive subgroup, the comparison between epochs highlighted BOLD signal variations concordant with the areas altered in the whole JME group. Indeed, in the IPS epochs associated with PPR, we observed a reduced IPS-induced activation of the visual cortex and a stronger deactivation of SM1 and the putamen, together with a more intense deactivation of DMN. These findings support the hypothesis of a shared pathophysiologic substrate between JME and photosensitivity. Dynamic evaluation of BOLD changes associated with PPR in photosensitive JME patients The dynamic analysis of PPR-related BOLD signal variations in the subgroup of photosensitive patients has provided us with further information on the role of the striato-thalamocortical network in the pathogenesis of JME. Previous work had investigated the time course of BOLD signal– related to spontaneous GSW, identifying an early thalamic activation,1 associated with a DMN deactivation and followed by a caudate nucleus deactivation.2 More recently Benuzzi et al.4 observed an early BOLD signal increase in the DMN, followed by thalamus activation. These results have highlighted the importance of subcortical structures in IGE although they cannot be directly compared to our data. Indeed we investigated hemodynamic phenomena related to PPR, which is a neurophysiological event other than spontaneous GSW. Only two studies had previously investigated photosensitivity by EEG-fMRI. In 1999 Hill et al.34 failed to detect any PPR-related BOLD signal changes. More recently Moeller et al.35 identified frontoparietal activation preceding PPR onset, followed by deactivation of the same regions. A comparison with our data is not straightforward. Both of these works investigated heterogeneous cohorts, whereas our study population was homogeneously constituted by people with JME. In our experiment positive BOLD changes were first detected in the putamen and in SM1. The early putamen activation confirms a major role of this structure in photosensitive JME. As reported earlier (see previous paragraph) structural and functional putamen abnormalities are acknowledged JME features. We hypothesize that the putamen could mediate between vision and motor-related areas, triggering a motor system hyperfunction corresponding to the PPR onset. We noticed that the putamen and SM1 are normally anticorrelated in healthy Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
1046 E. Bartolini et al. subjects, as positive BOLD changes in the former correspond to negative BOLD changes in the latter. In the dynamic analysis of PPR-related BOLD changes, we observed a hyper-correlation between putamen and primary sensorimotor area, as BOLD signal increased contemporarily in both of them. Together with the aforementioned activation, we observed BOLD signal decrease in DMN areas. Former studies have interpreted GSW-associated DMN deactivations as disturbance of the physiologic resting activity supporting the state of consciousness.3 A previously cited work has described DMN involvement in PPR occurrence in a heterogeneous group of photosensitive subjects, including one patient with JME.35 Considering PPR is less likely to be associated with disturbances of the level of consciousness, the interpretation of DMN deactivation remains a matter of debate. Impairment of DMN in people with JME has already been reported by Vollmar et al.,33 who suggested that an alteration of the DMN could lead to hyperconnectivity across systems, including the motor-related areas. In our study DMN showed an early BOLD decrease, at PPR onset, together with the putamen and SM1 activation. It is difficult to establish whether DMN could play a causal role or whether its changes reflect anticorrelation with SM1. After PPR onset we observed extended BOLD signal decrease involving formerly activated areas (i.e., putamen and SM1) and widely connected structures (i.e., thalamus and caudate nuclei), the interpretation of which is quite arduous. Of interest caudate nucleus deactivation has formerly been observed in studies on spontaneous GSW and interpreted as expression of a basal ganglia remote control system on absence seizures.2 The caudate nucleus is strongly involved in the motor circuitry and its deactivation could again point to a mechanism of PPR limitation. An analogous consideration could be proposed for BOLD signal reduction in the thalamus. As speculated by Moeller et al.,36 the thalamus could be fundamental for transition from subclinical PPR occurrence to generalized tonic–clonic seizures. Eventually the strong deactivation we observed could partly feature self-limiting mechanisms, in accordance with the fact that none of our patients had a seizure during the experiment. JME as a system epilepsy In this study IPS allowed us to disclose alterations of visual and motor circuits along with DMN in JME patients. These results reinforce the speculation that JME represents an example of system epilepsy, featuring the pathologic expression of a network made up of motor cortical areas and related cortical and subcortical structures.37 This hypothesis also finds support in recent structural studies, performed using diffusion tensor imaging (DTI). Different studies converge on microstructural abnormalities of white matter connecting cortical (especially Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
premotor-motor regions) and subcortical structures in JME29,38 and in photosensitive IGE.39 Alterations in fractional anisotropy or mean diffusivity of subcortical structures have been described in the thalamus38,39 and putamen.29 Limitations of the study Our results are consistent but may be partly biased by drug therapy. About half of patients were on valproate, which has been reported to interfere with basal ganglia circuitry.40 Therefore, we cannot rule out an influence on the caudate/putamen metabolism. Furthermore, the number of subjects in our study was relatively low; consequently, further studies performed on drug-naive patients and wider populations would be desirable. To assess whether our findings are JME-specific, we suggest investigations on different types of IGE and on photosensitive nonepileptic subjects.
Conclusion The results of our study are in agreement with the proposal of JME as a system epilepsy. People with JME exhibited an abnormal activation of the striato-thalamocortical network in response to IPS. The occurrence of PPR in photosensitive JME was associated with an early activation of the putamen and a hypercorrelation between the putamen itself and the primary sensorimotor area. We interpret massive deactivation in basal ganglia, thalami, and motorrelated areas as a possible self-limitation mechanism of epileptic discharges, but we cannot rule out it features a notspecific trait of the epilepsy phenotype. We propose that the PPR represents the final expression of pathogenic phenomena occurring in the striato-thalamocortical network, possibly core of the system epilepsy JME.
Acknowledgments The Authors thank V.K. Chinthapalli for his help in revising the manuscript; Annarita Ferrari, Melania Guida, Chiara Pizzanelli and Francesca Quaglia for the study population recruitment; and Rossella Buscemi and Silvia Tognazzi for their contribution in EEG preparation.
Disclosure None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
References 1. Hawco CS, Bagshaw AP, Lu Y, et al. BOLD changes occur prior to epileptic spikes seen on scalp EEG. Neuroimage 2007;35:1450–1458. 2. Moeller F, Siebner HR, Wolff S, et al. Changes in activity of striatothalamo-cortical network precede generalized spike wave discharges. Neuroimage 2008;39:1839–1849. 3. Vaudano AE, Laufs H, Kiebel SJ, et al. Causal hierarchy within the thalamo-cortical network in spike and wave discharges. PLoS ONE 2009;4:e6475.
1047 A Striato-Thalamocortical Hypothesis 4. Benuzzi F, Mirandola L, Pugnaghi M, et al. Increased cortical BOLD signal anticipates generalized spike and wave discharges in adolescents and adults with idiopathic generalized epilepsies. Epilepsia 2012;53:622–630. 5. Panayiotopoulos CP. Idiopathic generalized epilepsies: a review and modern approach. Epilepsia 2005;46(Suppl. 9):1–6. 6. Brigo F, Storti M, Benedetti MD, et al. Resting motor threshold in idiopathic generalized epilepsies: a systematic review with metaanalysis. Epilepsy Res 2012;101:3–13. 7. Groppa S, Siebner HR, Kurth C, et al. Abnormal response of motor cortex to photic stimulation in idiopathic generalized epilepsy. Epilepsia 2008;49:2022–2029. 8. Strigaro G, Prandi P, Varrasi C, et al. Intermittent photic stimulation affects motor cortex excitability in photosensitive idiopathic generalized epilepsy. Epilepsy Res 2013;104:78–83. 9. Siniatchkin M, Groppa S, Jerosch B, et al. Spreading photoparoxysmal EEG response is associated with an abnormal cortical excitability pattern. Brain 2007;130:78–87. 10. Porciatti V, Bonanni P, Fiorentini A, et al. Lack of cortical contrast gain control in human photosensitive epilepsy. Nat Neurosci 2000;3:259–263. 11. Takahashi Y, Fujiwara T, Yagi K, et al. Wavelength dependence of photoparoxysmal responses in photosensitive patients with epilepsy. Epilepsia 1999;40:23–27. 12. Wilkins AJ, Bonanni P, Porciatti V, et al. Physiology of photosensitivity. Epilepsia 2004;45:7–13. 13. Seneviratne U, Cook M, D’Souza W. The electroencephalogram of idiopathic generalized epilepsy. Epilepsia 2012;53:234–248. 14. Takahashi T, Nakasato N, Yokoyama H, et al. Low-luminance visual stimuli compared with stroboscopic IPS in eliciting PPR in photosensitive patients. Epilepsia 1999;40:44–49. 15. Commission on Classification and Terminology of the International League Against Epilepsy. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 1989;30:389–399. 16. Pesaresi I, Cosottini M, Belmonte G, et al. Reproducibility of BOLD localization of interictal activity in patients with focal epilepsy: intrasession and intersession comparisons. MAGMA 2011;24:285–296. 17. Waltz S, Christen HJ, Doose H. The different patterns of the photoparoxysmal response – a genetic study. Electroencephalogr Clin Neurophysiol 1992;83:138–145. 18. Jenkinson M, Bannister P, Brady M, et al. Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage 2002;17:825–841. 19. Mazziotta JC, Toga AW, Evans AC, et al. Digital brain atlases. Trends Neurosci 1995;18:210–211. 20. Beckmann CF, Smith S, Jenkinson M. General multi-level linear modelling for group analysis in fMRI. Neuroimage 2003;20:1052– 1063. 21. Agosta F, Pagani E, Rocca MA, et al. Voxel-based morphometry study of brain volumetry and diffusivity in amyotrophic lateral sclerosis patients with mild disability. Hum Brain Mapp 2007;28:1430–1438. 22. Takahashi T, Tsukahara Y. Pocket Monster incident and low luminance visual stimuli: special reference to deep red flicker stimulation. Acta Paediatr Jpn 1998;40:631–637. 23. Silva-Barrat C, Menini C, Bryere P, et al. Multiunitary activity analysis of cortical and subcortical structures in paroxysmal discharges and grand mal seizures in photosensitive baboons. Electroencephalogr Clin Neurophysiol 1986;64:455–468. 24. Strigaro G, Prandi P, Varrasi C, et al. Defective visual inhibition in photosensitive idiopathic generalized epilepsy. Epilepsia 2012;53:695–704. 25. Appleton R, Beirne M, Acomb B. Photosensitivity in juvenile myoclonic epilepsy. Seizure 2000;9:108–111.
26. Siniatchkin M, Moeller F, Shepherd A, et al. Altered cortical visual processing in individuals with a spreading photoparoxysmal EEG response. Eur J Neurosci 2007;26:529–536. 27. Vicente AF, Bermudez MA, Romero Mdel C, et al. Putamen neurons process both sensory and motor information during a complex task. Brain Res 2012;1466:70–81. 28. Vercueil L. Parkinsonism and epilepsy: case report and reappraisal of an old question. Epilepsy Behav 2000;1:128–130. 29. Keller SS, Ahrens T, Mohammadi S, et al. Microstructural and volumetric abnormalities of the putamen in juvenile myoclonic epilepsy. Epilepsia 2011;52:1715–1724. 30. Ciumas C, Wahlin TB, Jucaite A, et al. Reduced dopamine transporter binding in patients with juvenile myoclonic epilepsy. Neurology 2008;71:788–794. 31. Landvogt C, Buchholz HG, Bernedo V, et al. Alteration of dopamine D2/D3 receptor binding in patients with juvenile myoclonic epilepsy. Epilepsia 2010;51:1699–1706. 32. Uddin LQ, Kelly AM, Biswal BB, et al. Functional connectivity of default mode network components: correlation, anticorrelation, and causality. Hum Brain Mapp 2009;30:625–637. 33. Vollmar C, O’Muircheartaigh J, Symms MR, et al. Altered microstructural connectivity in juvenile myoclonic epilepsy: the missing link. Neurology 2012;78:1555–1559. 34. Hill RA, Chiappa KH, Huang-Hellinger F, et al. Hemodynamic and metabolic aspects of photosensitive epilepsy revealed by functional magnetic resonance imaging and magnetic resonance spectroscopy. Epilepsia 1999;40:912–920. 35. Moeller F, Siebner HR, Ahlgrimm N, et al. fMRI activation during spike and wave discharges evoked by photic stimulation. Neuroimage 2009;48:682–695. 36. Moeller F, Siebner HR, Wolff S, et al. Mapping brain activity on the verge of a photically induced generalized tonic-clonic seizure. Epilepsia 2009;50:1632–1637. 37. Avanzini G, Manganotti P, Meletti S, et al. The system epilepsies: a pathophysiological hypothesis. Epilepsia 2012;53:771–778. 38. Deppe M, Kellinghaus C, Duning T, et al. Nerve fiber impairment of anterior thalamocortical circuitry in juvenile myoclonic epilepsy. Neurology 2008;71:1981–1985. 39. Groppa S, Moeller F, Siebner H, et al. White matter microstructural changes of thalamocortical networks in photosensitivity and idiopathic generalized epilepsy. Epilepsia 2012;53:668–676. 40. Mahmoud F, Tampi RR. Valproic acid-induced parkinsonism in the elderly: a comprehensive review of the literature. Am J Geriatr Pharmacother 2011;9:405–412.
Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1. Nine gamma functions used as HRF to explore a time window starting 10 s before the PPR onset up until 10 s after that. Table S1. Demographic and clinical characteristics of the JME cohort. Table S2. Between-group comparisons of BOLD signal changes evoked by IPS: positive BOLD response. Table S3. Between-group comparisons of BOLD signal changes evoked by IPS: negative BOLD response.
Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
Abnormal response to photic stimulation in Juvenile Myoclonic Epilepsy: An EEG-fMRI study *Emanuele Bartolini, †Ilaria Pesaresi, *Serena Fabbri, *Paolo Cecchi, ‡Filippo Sean Giorgi, *Ferdinando Sartucci, *Ubaldo Bonuccelli, and §Mirco Cosottini Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
SUMMARY
Emanuele Bartolini is a Research Fellow at the University of Pisa, Italy.
Objective: Juvenile myoclonic epilepsy (JME) is a young-onset electroclinical syndrome, characterized by myoclonic, generalized tonic–clonic, and possibly typical absence seizures. Interictal electroencephalography (EEG) displays 3–6 Hz spike/ polyspike and wave pattern. Photosensitivity is common. Our aim was to explore the blood oxygen level–dependent (BOLD) response evoked by a highly provocative photic stimulus in a cohort of people with JME compared to a group of nonphotosensitive healthy controls, and to investigate the hemodynamic phenomena seen in patients with photosensitive JME. Methods: We studied 13 JME patients and 18 healthy controls using EEG–functional magnetic resonance imaging (fMRI) performed during low luminance intermittent photic stimulation (IPS). The BOLD response to IPS was investigated both in JME and control groups. In photosensitive JME subjects, we also performed a dynamic evaluation of BOLD signal changes evoked by the photoparoxysmal response (PPR) in a time frame ranging from 10 s before the onset of the EEG paroxysm up until 10 s afterward. Results: The IPS evoked a positive BOLD response in striate and extrastriate visual areas, which was less in JME patients than in controls. Moreover, people with JME had a reduced positive BOLD response in the frontoparietal areas and putamen but a stronger negative BOLD response in the primary sensorimotor cortex (SM1) and in cortical regions belonging to the default mode network (DMN). In JME, the dynamic evaluation of BOLD signal changes related to PPR revealed an early positive response in the putamen and SM1, followed by BOLD signal decrements in the putamen, caudate nuclei, thalami, and SM1. Significance: Our results confirm the hypothesis that people with JME might have an altered interaction between the motor circuit and other neuronal networks, with prominent involvement of basal ganglia circuitry. The PPR could be a final expression of pathogenic phenomena occurring in the striato-thalamocortical system, possibly a core feature of system epilepsy JME. KEY WORDS: Juvenile Myoclonic Epilepsy, Neuroimaging, Neurophysiology, Photosensitivity.
Accepted March 19, 2014; Early View publication May 23, 2014. *Department of Clinical and Experimental Medicine, University of Pisa, Italy; †Unit of Neuroradiology, Santa Chiara Hospital AOUP, Italy; ‡Unit of Neurology, Santa Chiara Hospital AOUP, Pisa, Italy; and §Department of Translational Research and New Technologies in Medicine and Surgery, University of Pisa, Pisa, Italy Address correspondence to Mirco Cosottini, Department of Translational Research and New Technologies in Medicine and Surgery, University of Pisa, Via Risorgimento 36, 56126 Pisa, Italy. E-mail: mircocosottini @libero.it Wiley Periodicals, Inc. © 2014 International League Against Epilepsy
The recent extensive application of electroencephalography–functional magnetic resonance imaging (EEG-fMRI) in idiopathic generalized epilepsy (IGE) has challenged the common interpretation of syndromes featuring primarily diffuse cortical discharges, and suggests the involvement of complex neuronal networks.1–4 The majority of these articles have addressed the whole IGE group or subgroup constituted by electroclinical phenotypes (e.g., spontaneous generalized spike and waves [GSWs] or absence seizures
1038
1039 A Striato-Thalamocortical Hypothesis [ABS]). No EEG-fMRI study has specifically explored the most diffuse form of IGE, juvenile myoclonic epilepsy (JME). This is a genetically determined syndrome characterized by adolescent/adult-onset myoclonic jerks alone or combined with generalized tonic–clonic seizures (GTCS) and ABS. Interictal EEG is typically characterized by an irregular mixture of 3–6 Hz spike/polyspike and slow wave activity.5 From the neurophysiological point of view, JME is characterized by a lower resting motor threshold than normal and by stronger intracortical c-aminobutyric acid (GABA)ergic inhibition of the motor system in response to photic stimulation.6–8 The latter characteristic is shared with healthy photosensitive subjects.9 Photosensitivity is characterized by abnormalities in the visual system (e.g., altered visual contrast gain control, heightened sensitivity to specific wavelengths).10,11 Whenever a luminous stimulus is sufficiently provocative, a critical mass of occipital neurons are activated, yielding epileptiform activity that may spread toward different areas, especially the frontal lobes.12 JME is deemed to be the most photosensitive form of epilepsy. About one third of patients with JME have a photoparoxysmal response (PPR) during classic stroboscopic intermittent photic stimulation (IPS).13 The ratio could even be higher with more provocative stimulation paradigms, such as low luminance ( 600 nm) flickers.14 We applied EEG-fMRI to study hemodynamic cortical and subcortical response to IPS in a cohort of people with JME compared to a group of healthy subjects. EEG-fMRI is a methodology based on synchronous registration of fMRI and EEG data. It can be used to detect changes in the blood oxygen level–dependent (BOLD) response to external or neurophysiologic events. In this context IPS and PPR occurrence represent phenomena that can be triggered during fMRI, offering a privileged time window to study the associated hemodynamic phenomena. We hypothesize that JME is characterized by a cortical and subcortical network disruption that predisposes to photosensitivity. Starting from this consideration, we investigated the pattern of BOLD response evoked by a highly provocative photic stimulation (deep red low luminance flickers) in people with JME. Furthermore, we aimed to investigate the hemodynamic phenomena that precede and accompany the occurrence of PPR in photosensitive JME patients.
Methods Subjects We enrolled 13 JME patients, selected consecutively from the Epilepsy Clinic of the University of Pisa (mean age 29.3 years, range 16–49 years). Inclusion criteria were the following: (1) definite diagnosis of JME according to the International League Against Epilepsy (ILAE) classification,15 (2) unremarkable findings on conventional brain
MRI scan, and (3) routine EEG with a normal background. Patients’ clinical histories were characterized by myoclonic seizures (n = 13/13), GTCS (n = 12/13), ABS (n = 5/13), and at least one interictal EEG with generalized spike-polyspike and wave (n = 13/13). Demographic and clinical characteristics of the JME cohort are summarized in Table S1. We also recruited 18 healthy controls with no history of febrile convulsions, seizures, or family history for epilepsy. Three healthy controls were excluded for PPR recording during the scan. Included controls were matched with patients for sex (12 female, three male) and age (mean age 26.4 years; range 19–34 years; t-test p-value = 0.2226). The study was approved by the local ethics committee. Written informed consent was obtained from each subject participating in the study, or by the patient’s parents for participants who were younger than 18 years. EEG acquisition, processing, and analysis EEG was continuously recorded within the scanner by using an MRI-compatible cap with 32-electrodes and an MRI-compatible amplifier (BrainAmp-MR; Brain Products, Gilching, Germany). The EEG acquisition and processing is described in detail elsewhere.16 EEG recordings were reviewed offline by an experienced electroencephalographer (E.B.) for identification of EEG abnormalities related to IPS. We analyzed only diffuse PPR, classified as class 4 according to Waltz et al.17 Patients who presented with at least one PPR during the experiment were classified as photosensitive. Photic stimulation inside the scanner In the fMRI experiment IPS was administered in a block design. According to previous work,14 IPS consisted of deep red (wave length 680 nm) low luminance (luminance < 30 cd/m2) flashes of different frequencies (10, 15, and 20 Hz), using 20 s blocks of stimulation alternating with 15 s blocks of rest. IPS blocks of different frequencies were presented in pseudo-random order (every frequency was repeated five times). The subjects were instructed to keep eyes open during EEG-fMRI examination. The light was delivered by an MRI compatible visual stimulation device (VisuaStimResonance Technologies, Northridge, CA, U.S.A.) composed of a fiberoptic goggle system, playing as a liquid-crystal display (LCD) screen placed in front of the eyes. The visual paradigm was managed by the software Presentation (Neurobehavioral Systems, Berkeley, CA, U.S.A.). The EEG was observed online to interrupt the stimulation in case of a seizure or a prolonged PPR. Subjects were instructed to push an emergency button in case of discomfort during IPS. fMRI acquisition fMRI data were acquired continuously during EEG recordings with a 3T scanner (Discovery MR 750; GE Medical System, Milwaukee, WI, U.S.A.). The fMRI data were obtained by T2*-weighted gradient-recalled echo-planarEpilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
1040 E. Bartolini et al. imaging (EPI) sequence (TR 2,500 msec, TE 40 msec, FA 90 degrees, image matrix 128 9 128, in-plane field-of-view 260 9 260 mm2) with 28 interleaved slices (slice thickness 4 mm, gap 1 mm) angled at 30 degrees from the anterior– posterior commissural plane, repeated over 210 volumes for a total scanning time of 8 min 30 s. For each patient, additional high-resolution T1-weighted images were acquired (three dimensional (3D)-BRAVO sequence, TR 8.16 msec, TI 450 msec, TE 3.18 msec, FA 12 degrees, voxel size 1 9 1 9 1 mm, 160 axial-oblique slices, total scanning time 4 min 28 s) to provide accurate anatomic references for functional data.
Both positive and negative BOLD changes were investigated in first-level and high-level statistical analysis.
fMRI data processing Data were analyzed by using fMRI Expert Analysis Tool (FEAT) v5.98 tool of the software-package FSL 4.1.5 (http://www.fmrib.ox.ac.uk/fsl/). Both patient and control datasets underwent the same analysis steps. Each functional dataset underwent preliminary processing including slicescan-time-correction, 3D motion-correction, high-passtemporal-filtering (100 s) and spatial-smoothing (Gaussian Kernel, 8 mm full-width-half-maximum). All subjects showed movement-related displacement lower than 3 mm/ 3 degrees.
Dynamic evaluation of BOLD changes associated with PPR in photosensitive JME patients In photosensitive patients we explored the temporal dynamic of BOLD signal changes associated with PPR by an event-related design. The PPR regressor was modeled with onset and duration of the events and convolved with a gamma-HRF. We explored a time window ranging from 10 s before the onset of the EEG paroxysm up until 10 s afterward, by a set of nine GLMs with PPR regressors shifted by 2.5 s in time (Fig. S1). Each GLM also included an IPS regressor (a unique regressor including all the photic stimulation frequencies) and motion displacement parameters as confounds. Group analysis was performed by a fixed effect model to obtain a mean group map of PPR regressor at each temporal bin. In total nine mean group positive BOLD maps and nine mean group negative BOLD maps were obtained. A statistical Z-threshold of 4.0 and cluster p-threshold < 0.05 were applied to all of them.
Comparison of BOLD response evoked by IPS in JME patients and healthy controls First-level statistical analysis was performed by the general-linear-model (GLM) approach. The signal model in each voxel was derived by block-design paradigm with three distinct predictors for each IPS frequency (10, 15, and 20 Hz). The expected BOLD responses were obtained by convolving the original predictor waveforms with a gamma-form hemodynamic response function (standard HRF peak = 6 s). The design matrix of GLM also included motion displacement parameters estimated by Motion Correction using FMRIB's Linear Image Registration Tool (MCFLIRT) as confound variables. For patients with PPR, a dedicated regressor was utilized and added to GLM. To obtain spatially resolved anatomic references for statistical maps, functional datasets were registered to brain-extracted 3D T1-weighted images by linear transformations with 6-degrees-of-freedom (FLIRT tool).18 Then functional and structural datasets were aligned to Montreal Neurological Institute - 152 (MNI-152) standard space (International Consortium for Brain Mapping-152 [ICBM-152] template) by linear 12-degrees-of-freedom transformations (FLIRT tool).19 The higher-level statistical analysis was performed by the fixed-effect model.20 The Z statistical maps resulting from both intragroup analysis and between-group analysis underwent cluster thresholding (Z-threshold = 2.0, cluster pthreshold < 0.05). For between-group statistical analysis in subcortical nuclei, a small volume correction was applied to correct for multiple comparisons (Z-threshold = 2.0, cluster p-threshold < 0.05).21 Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
Comparison of IPS epochs with PPR and without PPR in photosensitive JME patients In photosensitive patients we compared the BOLD signal changes during IPS epochs associated with PPR to epochs without PPR. By detail, IPS blocks with and without PPR were used to build separate regressors in GLM. Both regressors were convolved with gamma-HRF. High-level intragroup analysis was performed by a fixed effect model (Z-threshold = 2.0, cluster p-threshold < 0.05).
Results During EEG-fMRI examinations, IPS evoked at least one PPR in 7 of 13 patients (total number of PPR recorded: 41; mean event duration 1.27 s, range 0.20–4.32 s; Table S1). None of the patients or controls reported discomfort during the experiment. Comparison of BOLD response evoked by IPS in JME patients and healthy controls In healthy subjects, IPS evoked a positive BOLD response in striate and extrastriate visual areas at all frequencies. A positive BOLD response was also observed in frontoparietal areas and in the putamen (especially elicited by 20 Hz IPS). In addition, negative BOLD changes were induced in the primary sensorimotor area (SM1), frontal, parietal, temporal and cingulate cortices (Fig. 1). The pattern of BOLD decrements in frontoparietal areas, temporal cortices, and cingulate cortices resembled the spatial distribution of the so-called default-mode–network (DMN). Patients showed a significantly lower positive BOLD response compared to controls in striate and extrastriate
1041 A Striato-Thalamocortical Hypothesis
Figure 1. BOLD changes evoked by low luminance intermittent photic stimulation in healthy controls. Positive (red-yellow) and negative (bluegreen) BOLD responses evoked by IPS in healthy controls. Epilepsia ILAE
Figure 2. BOLD changes evoked by low luminance intermittent photic stimulation in patients with JME. Positive (red-yellow) and negative (bluegreen) BOLD responses evoked by IPS in patients with JME. Epilepsia ILAE Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
1042 E. Bartolini et al.
Figure 3. Comparison between BOLD changes evoked by low luminance IPS in healthy controls and patients with JME. On the left side: betweengroup differences in positive (red-yellow) BOLD response evoked by IPS. On the right side: between-group differences in negative (bluegreen) BOLD response evoked by IPS. Epilepsia ILAE
visual areas (Brodmann area BA17, 18, 19), in parietal (BA5, 7, 39) and frontal areas (BA6, 9, 44, 46) and in putamen (Figs. 2 and 3 and Table S2). A stronger negativeBOLD response in JME was observed in SM1 (BA1, 2, 3, 4), premotor cortex (BA6), and in the DMN areas (parietal cortex BA39; temporal cortex BA20, 21, 22; frontal cortex BA8, 9, 44, 46; cingulate cortex (BA23, 24, 30, 32; Figs. 2 and 3 and Table S3). Comparison of IPS epochs with PPR and without PPR in photosensitive JME patients In photosensitive patients, IPS epochs associated with PPR were characterized by a weaker positive BOLD response in the striate and extrastriate cortex compared to IPS epochs without PPR. In addition, IPS with PPR elicited a stronger negative BOLD response in SM1, the premotor cortex, DMN, and the putamen (Fig. 4). Dynamic evaluation of BOLD changes associated with PPR in photosensitive JME patients The dynamic analysis of PPR events disclosed an early positive BOLD response in putamen and SM1 accompanied by a negative BOLD response in DMN (0 and +2.5 s from Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
PPR onset). Then, positive BOLD phenomena disappeared and negative BOLD changes surged in the putamen, the head of the caudate nuclei, and the frontomesial cortex (+5 s from the PPR onset). Further on, negative BOLD changes involved the thalami and SM1 (+7.5 s from PPR onset) and then progressively extended to broad cortical regions (+10 s from PPR onset) (Fig. 5).
Discussion The aim of our study was investigating JME pathophysiology by addressing the hemodynamic effects on cortical and subcortical brain structures induced by highly provocative photic stimuli. To build an MRI-compatible IPS paradigm with the maximal epileptogenic potential, we applied photic stimulation with trains of low luminance deep red flickers, generated outside the magnet room and driven to the patient via fibre optics and goggles. Low luminance IPS was studied extensively after the “Pocket Monster incident,” when about 600 Japanese children had seizure while watching a cartoon on a TV screen. The provocative event was identified as a sequence of deep red flickers (wave length >600 nm).22
1043 A Striato-Thalamocortical Hypothesis
A
B
C
D
Figure 4. Comparison between IPS epochs with PPR and without PPR. (A) Positive (red-yellow) BOLD responses induced by IPS epochs without PPR (upper row) and with PPR (lower row). (B) Negative (blue-green) BOLD responses elicited by IPS epochs without PPR (upper row) and with PPR (lower row). (C, D) Differences between BOLD signal changes evoked by IPS epochs with and without PPR (on the left side positive BOLD changes, red-yellow; on the right side negative BOLD changes, blue-green). Epilepsia ILAE
Subsequently Takahashi et al.14 demonstrated that deep red low luminance stimulation is more provocative than conventional stroboscopic IPS performed by high luminance flashes. Photosensitivity is generally regarded as a cortically driven phenomenon, mainly involving the occipital lobes, with a secondary contribution from the frontal areas.23 The development of a PPR probably occurs when normal physiologic excitation of the visual cortex exceeds a critical level as a result of a failure in inhibitory processes. The latter would be characteristic of photosensitive people, possibly as a genetic trait. Consistent with this hypothesis, Strigaro et al.24 have recently demonstrated that photosensitive IGE is characterized by a flaw in the inhibition of the visual system normally induced by a single luminous flash. Previous work has described complementary abnormalities in the visual system of people with photosensitive epilepsy. Porciatti et al.10 reported a defective gain control mechanism of visual stimuli in photosensitive epilepsy, possibly related to defective inhibition, highlighted by visual evoked potentials (VEPs). Takahashi et al. proposed two different pathophysiologic mechanisms of photosensitivity: a “quantity of light dependent” and a “wavelength dependent” mechanism. The first would require stimuli with an elevated
amount of luminance contrast to elicit a PPR; the second would be based on the high provocative properties of low luminance flashes with wavelengths of around 700 nm (i.e., deep red), and would be more often observed in people with IGE and photosensitive healthy subjects.11 The paradigm of stimulation we applied acts by the “wavelength dependent” mechanism. In detail, the color theory states that human visual system carries information about color through antagonistic channels along the parvocellular and magnocellular pathways. These channels derive from the overlapping inputs of three retinal cones (L-cones for red, M-cones for green, S-cones for blue). It has been speculated that low luminance deep red stimuli activate only retinal red cones (L-cones) with no effect on counteracting antagonist cones (M-cones and S-cones), yielding the maximal provocative effect.12 Our stimulus protocol was reinforced by the use of long-lasting stimulation trains, much longer than in former studies.14,22 Elongation of the stimulus train was established because of the previous observation that IPS duration increases the rate of photosensitivity in people with JME.25 We applied three randomordered frequencies, acknowledged as highly provocative (i.e., 10, 15, and 20 Hz).25 The strong provocative effect of Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
1044 E. Bartolini et al.
Figure 5. Dynamic evaluation of BOLD changes associated with PPR in photosensitive JME patients. Positive (red-yellow) and negative (blue-green) BOLD responses related to PPR in JME patients across a 20 s time window (from 10 s before PPR onset up till 10 s after that). Upper rows: axial view at level of basal ganglia. Lower rows: upper view of the cerebral volume rendering. Epilepsia ILAE
the visual stimulation we applied was in fact highlighted by the occurrence of PPR in 3 of 18 healthy subjects. BOLD response evoked by IPS in JME patients and healthy controls People with JME were characterized by a reduced BOLD response to IPS in the striate and extrastriate visual cortex compared to healthy controls. Up until now no definite data have been available on the visual cortex excitability in JME subjects. Previous transcranial magnetic stimulation (TMS) studies demonstrated a heightened excitability of the primary visual area to photosensitivity. In fact, the presence of PPR in healthy subjects is associated with lower phosphene thresholds and steeper stimuEpilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
lus-response curves in studies performed by TMS stimulation over the primary visual cortex.9 Hyperexcitability would affect both magnocellular and parvocellular systems, as demonstrated by pattern-reverse VEPs.26 Furthermore, photosensitive subjects are characterized by stronger habituation of the visual cortex to trains of stimuli, as shown by a stronger inhibitory effect of TMS on visual perception and by a more pronounced VEP habituation to pattern-reverse blocks.9,26 The reduced IPS-evoked positive BOLD response that we observed in patients’ visual cortex could be explained by IPS repetitive nature. Similar to in healthy photosensitive subjects, we hypothesize people with JME could be characterized by a stronger habituation to a repetitive stimulus.
1045 A Striato-Thalamocortical Hypothesis Along with a reduced activation of striate and extrastriate occipital visual cortex, patients presented with a lower BOLD response than controls in vision-related frontoparietal areas. In addition, photic stimulation provided data on motor circuits, unveiling an altered BOLD response in motorrelated cortical and subcortical areas. In detail, a stronger BOLD signal decrease was observed in the SM1 and premotor cortex of patients compared to controls, with maximum effect at 10 Hz frequency. The interpretation of such results is not straightforward. Studies performed by TMS have previously demonstrated that photic stimulation normally increases the motor cortex excitability, thereby shortening the cortical silent period (CSP) in healthy not-photosensitive subjects. On the contrary, IPS does not modify CSP either in healthy photosensitive subjects or in people with IGE/JME (whether photosensitive or not).7–9 CSP is a measure of inhibitory processes within the motor cortex, mediated by GABA neurons. Resting motor threshold (rMT) is instead believed to reflect membrane excitability of corticospinal neurons and related interneurons within the motor cortex. Siniatchkin et al.9 studied photosensitive and nonphotosensitive healthy subjects and did not detect any rMT differences between the groups. Consequently we may speculate that the differences in rMT can be attributed to factors other than the photosensitive trait. Of interest a recent meta-analysis by Brigo et al.6 has assessed that drug-naive people with JME have a lower rMT and a heightened motor excitability compared to healthy subjects. Hence, the sensorimotor cortex of people with JME would be characterized by a major excitability balanced by stronger GABAergic intracortical inhibition. The stronger BOLD deactivation in SM1 that we detected in patients could express a more intense GABAergic intracortical inhibition. With regard to subcortical structures, IPS induced a positive BOLD response in the putamen nuclei of healthy controls, which was lacking in patients. A role of the putamen in visual-motor integration has formerly been hypothesized, with the description of different classes of neurons responding specifically to pure visual stimuli next to neurons involved in motor planning/inhibition.27 Dysfunction of basal ganglia circuitry has already been reported in epilepsy.28 Specifically JME has been associated with microstructural and macrostructural abnormalities in the putamen29 and with dopaminergic dysfunction. In particular, PET studies have demonstrated a reduced binding to the dopamine transporter in the substantia nigra and midbrain and reduced binding to dopamine D2/D3 receptors in the posterior putamen.30,31 An abnormal BOLD response to IPS was also observed in areas belonging to the DMN, in particular a stronger BOLD signal reduction was detected in JME patients compared to controls. Normally DMN is anticorrelated to the sensorimotor network.32 On the contrary we observed people with JME
had deactivation in both systems. A previous study has already described an alteration of DMN in JME patients, suggesting that DMN impairment could lead to hyperconnectivity across systems, including the motor-related areas.33 Because DMN abnormalities have been reported in many different epileptic phenotypes independently of photosensitivity,2,4,33 it is difficult to establish whether DMN could play a causal role or it represents a not specific trait. Comparison of IPS epochs with PPR and without PPR in photosensitive JME patients In the photosensitive subgroup, the comparison between epochs highlighted BOLD signal variations concordant with the areas altered in the whole JME group. Indeed, in the IPS epochs associated with PPR, we observed a reduced IPS-induced activation of the visual cortex and a stronger deactivation of SM1 and the putamen, together with a more intense deactivation of DMN. These findings support the hypothesis of a shared pathophysiologic substrate between JME and photosensitivity. Dynamic evaluation of BOLD changes associated with PPR in photosensitive JME patients The dynamic analysis of PPR-related BOLD signal variations in the subgroup of photosensitive patients has provided us with further information on the role of the striato-thalamocortical network in the pathogenesis of JME. Previous work had investigated the time course of BOLD signal– related to spontaneous GSW, identifying an early thalamic activation,1 associated with a DMN deactivation and followed by a caudate nucleus deactivation.2 More recently Benuzzi et al.4 observed an early BOLD signal increase in the DMN, followed by thalamus activation. These results have highlighted the importance of subcortical structures in IGE although they cannot be directly compared to our data. Indeed we investigated hemodynamic phenomena related to PPR, which is a neurophysiological event other than spontaneous GSW. Only two studies had previously investigated photosensitivity by EEG-fMRI. In 1999 Hill et al.34 failed to detect any PPR-related BOLD signal changes. More recently Moeller et al.35 identified frontoparietal activation preceding PPR onset, followed by deactivation of the same regions. A comparison with our data is not straightforward. Both of these works investigated heterogeneous cohorts, whereas our study population was homogeneously constituted by people with JME. In our experiment positive BOLD changes were first detected in the putamen and in SM1. The early putamen activation confirms a major role of this structure in photosensitive JME. As reported earlier (see previous paragraph) structural and functional putamen abnormalities are acknowledged JME features. We hypothesize that the putamen could mediate between vision and motor-related areas, triggering a motor system hyperfunction corresponding to the PPR onset. We noticed that the putamen and SM1 are normally anticorrelated in healthy Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
1046 E. Bartolini et al. subjects, as positive BOLD changes in the former correspond to negative BOLD changes in the latter. In the dynamic analysis of PPR-related BOLD changes, we observed a hyper-correlation between putamen and primary sensorimotor area, as BOLD signal increased contemporarily in both of them. Together with the aforementioned activation, we observed BOLD signal decrease in DMN areas. Former studies have interpreted GSW-associated DMN deactivations as disturbance of the physiologic resting activity supporting the state of consciousness.3 A previously cited work has described DMN involvement in PPR occurrence in a heterogeneous group of photosensitive subjects, including one patient with JME.35 Considering PPR is less likely to be associated with disturbances of the level of consciousness, the interpretation of DMN deactivation remains a matter of debate. Impairment of DMN in people with JME has already been reported by Vollmar et al.,33 who suggested that an alteration of the DMN could lead to hyperconnectivity across systems, including the motor-related areas. In our study DMN showed an early BOLD decrease, at PPR onset, together with the putamen and SM1 activation. It is difficult to establish whether DMN could play a causal role or whether its changes reflect anticorrelation with SM1. After PPR onset we observed extended BOLD signal decrease involving formerly activated areas (i.e., putamen and SM1) and widely connected structures (i.e., thalamus and caudate nuclei), the interpretation of which is quite arduous. Of interest caudate nucleus deactivation has formerly been observed in studies on spontaneous GSW and interpreted as expression of a basal ganglia remote control system on absence seizures.2 The caudate nucleus is strongly involved in the motor circuitry and its deactivation could again point to a mechanism of PPR limitation. An analogous consideration could be proposed for BOLD signal reduction in the thalamus. As speculated by Moeller et al.,36 the thalamus could be fundamental for transition from subclinical PPR occurrence to generalized tonic–clonic seizures. Eventually the strong deactivation we observed could partly feature self-limiting mechanisms, in accordance with the fact that none of our patients had a seizure during the experiment. JME as a system epilepsy In this study IPS allowed us to disclose alterations of visual and motor circuits along with DMN in JME patients. These results reinforce the speculation that JME represents an example of system epilepsy, featuring the pathologic expression of a network made up of motor cortical areas and related cortical and subcortical structures.37 This hypothesis also finds support in recent structural studies, performed using diffusion tensor imaging (DTI). Different studies converge on microstructural abnormalities of white matter connecting cortical (especially Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
premotor-motor regions) and subcortical structures in JME29,38 and in photosensitive IGE.39 Alterations in fractional anisotropy or mean diffusivity of subcortical structures have been described in the thalamus38,39 and putamen.29 Limitations of the study Our results are consistent but may be partly biased by drug therapy. About half of patients were on valproate, which has been reported to interfere with basal ganglia circuitry.40 Therefore, we cannot rule out an influence on the caudate/putamen metabolism. Furthermore, the number of subjects in our study was relatively low; consequently, further studies performed on drug-naive patients and wider populations would be desirable. To assess whether our findings are JME-specific, we suggest investigations on different types of IGE and on photosensitive nonepileptic subjects.
Conclusion The results of our study are in agreement with the proposal of JME as a system epilepsy. People with JME exhibited an abnormal activation of the striato-thalamocortical network in response to IPS. The occurrence of PPR in photosensitive JME was associated with an early activation of the putamen and a hypercorrelation between the putamen itself and the primary sensorimotor area. We interpret massive deactivation in basal ganglia, thalami, and motorrelated areas as a possible self-limitation mechanism of epileptic discharges, but we cannot rule out it features a notspecific trait of the epilepsy phenotype. We propose that the PPR represents the final expression of pathogenic phenomena occurring in the striato-thalamocortical network, possibly core of the system epilepsy JME.
Acknowledgments The Authors thank V.K. Chinthapalli for his help in revising the manuscript; Annarita Ferrari, Melania Guida, Chiara Pizzanelli and Francesca Quaglia for the study population recruitment; and Rossella Buscemi and Silvia Tognazzi for their contribution in EEG preparation.
Disclosure None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
References 1. Hawco CS, Bagshaw AP, Lu Y, et al. BOLD changes occur prior to epileptic spikes seen on scalp EEG. Neuroimage 2007;35:1450–1458. 2. Moeller F, Siebner HR, Wolff S, et al. Changes in activity of striatothalamo-cortical network precede generalized spike wave discharges. Neuroimage 2008;39:1839–1849. 3. Vaudano AE, Laufs H, Kiebel SJ, et al. Causal hierarchy within the thalamo-cortical network in spike and wave discharges. PLoS ONE 2009;4:e6475.
1047 A Striato-Thalamocortical Hypothesis 4. Benuzzi F, Mirandola L, Pugnaghi M, et al. Increased cortical BOLD signal anticipates generalized spike and wave discharges in adolescents and adults with idiopathic generalized epilepsies. Epilepsia 2012;53:622–630. 5. Panayiotopoulos CP. Idiopathic generalized epilepsies: a review and modern approach. Epilepsia 2005;46(Suppl. 9):1–6. 6. Brigo F, Storti M, Benedetti MD, et al. Resting motor threshold in idiopathic generalized epilepsies: a systematic review with metaanalysis. Epilepsy Res 2012;101:3–13. 7. Groppa S, Siebner HR, Kurth C, et al. Abnormal response of motor cortex to photic stimulation in idiopathic generalized epilepsy. Epilepsia 2008;49:2022–2029. 8. Strigaro G, Prandi P, Varrasi C, et al. Intermittent photic stimulation affects motor cortex excitability in photosensitive idiopathic generalized epilepsy. Epilepsy Res 2013;104:78–83. 9. Siniatchkin M, Groppa S, Jerosch B, et al. Spreading photoparoxysmal EEG response is associated with an abnormal cortical excitability pattern. Brain 2007;130:78–87. 10. Porciatti V, Bonanni P, Fiorentini A, et al. Lack of cortical contrast gain control in human photosensitive epilepsy. Nat Neurosci 2000;3:259–263. 11. Takahashi Y, Fujiwara T, Yagi K, et al. Wavelength dependence of photoparoxysmal responses in photosensitive patients with epilepsy. Epilepsia 1999;40:23–27. 12. Wilkins AJ, Bonanni P, Porciatti V, et al. Physiology of photosensitivity. Epilepsia 2004;45:7–13. 13. Seneviratne U, Cook M, D’Souza W. The electroencephalogram of idiopathic generalized epilepsy. Epilepsia 2012;53:234–248. 14. Takahashi T, Nakasato N, Yokoyama H, et al. Low-luminance visual stimuli compared with stroboscopic IPS in eliciting PPR in photosensitive patients. Epilepsia 1999;40:44–49. 15. Commission on Classification and Terminology of the International League Against Epilepsy. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 1989;30:389–399. 16. Pesaresi I, Cosottini M, Belmonte G, et al. Reproducibility of BOLD localization of interictal activity in patients with focal epilepsy: intrasession and intersession comparisons. MAGMA 2011;24:285–296. 17. Waltz S, Christen HJ, Doose H. The different patterns of the photoparoxysmal response – a genetic study. Electroencephalogr Clin Neurophysiol 1992;83:138–145. 18. Jenkinson M, Bannister P, Brady M, et al. Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage 2002;17:825–841. 19. Mazziotta JC, Toga AW, Evans AC, et al. Digital brain atlases. Trends Neurosci 1995;18:210–211. 20. Beckmann CF, Smith S, Jenkinson M. General multi-level linear modelling for group analysis in fMRI. Neuroimage 2003;20:1052– 1063. 21. Agosta F, Pagani E, Rocca MA, et al. Voxel-based morphometry study of brain volumetry and diffusivity in amyotrophic lateral sclerosis patients with mild disability. Hum Brain Mapp 2007;28:1430–1438. 22. Takahashi T, Tsukahara Y. Pocket Monster incident and low luminance visual stimuli: special reference to deep red flicker stimulation. Acta Paediatr Jpn 1998;40:631–637. 23. Silva-Barrat C, Menini C, Bryere P, et al. Multiunitary activity analysis of cortical and subcortical structures in paroxysmal discharges and grand mal seizures in photosensitive baboons. Electroencephalogr Clin Neurophysiol 1986;64:455–468. 24. Strigaro G, Prandi P, Varrasi C, et al. Defective visual inhibition in photosensitive idiopathic generalized epilepsy. Epilepsia 2012;53:695–704. 25. Appleton R, Beirne M, Acomb B. Photosensitivity in juvenile myoclonic epilepsy. Seizure 2000;9:108–111.
26. Siniatchkin M, Moeller F, Shepherd A, et al. Altered cortical visual processing in individuals with a spreading photoparoxysmal EEG response. Eur J Neurosci 2007;26:529–536. 27. Vicente AF, Bermudez MA, Romero Mdel C, et al. Putamen neurons process both sensory and motor information during a complex task. Brain Res 2012;1466:70–81. 28. Vercueil L. Parkinsonism and epilepsy: case report and reappraisal of an old question. Epilepsy Behav 2000;1:128–130. 29. Keller SS, Ahrens T, Mohammadi S, et al. Microstructural and volumetric abnormalities of the putamen in juvenile myoclonic epilepsy. Epilepsia 2011;52:1715–1724. 30. Ciumas C, Wahlin TB, Jucaite A, et al. Reduced dopamine transporter binding in patients with juvenile myoclonic epilepsy. Neurology 2008;71:788–794. 31. Landvogt C, Buchholz HG, Bernedo V, et al. Alteration of dopamine D2/D3 receptor binding in patients with juvenile myoclonic epilepsy. Epilepsia 2010;51:1699–1706. 32. Uddin LQ, Kelly AM, Biswal BB, et al. Functional connectivity of default mode network components: correlation, anticorrelation, and causality. Hum Brain Mapp 2009;30:625–637. 33. Vollmar C, O’Muircheartaigh J, Symms MR, et al. Altered microstructural connectivity in juvenile myoclonic epilepsy: the missing link. Neurology 2012;78:1555–1559. 34. Hill RA, Chiappa KH, Huang-Hellinger F, et al. Hemodynamic and metabolic aspects of photosensitive epilepsy revealed by functional magnetic resonance imaging and magnetic resonance spectroscopy. Epilepsia 1999;40:912–920. 35. Moeller F, Siebner HR, Ahlgrimm N, et al. fMRI activation during spike and wave discharges evoked by photic stimulation. Neuroimage 2009;48:682–695. 36. Moeller F, Siebner HR, Wolff S, et al. Mapping brain activity on the verge of a photically induced generalized tonic-clonic seizure. Epilepsia 2009;50:1632–1637. 37. Avanzini G, Manganotti P, Meletti S, et al. The system epilepsies: a pathophysiological hypothesis. Epilepsia 2012;53:771–778. 38. Deppe M, Kellinghaus C, Duning T, et al. Nerve fiber impairment of anterior thalamocortical circuitry in juvenile myoclonic epilepsy. Neurology 2008;71:1981–1985. 39. Groppa S, Moeller F, Siebner H, et al. White matter microstructural changes of thalamocortical networks in photosensitivity and idiopathic generalized epilepsy. Epilepsia 2012;53:668–676. 40. Mahmoud F, Tampi RR. Valproic acid-induced parkinsonism in the elderly: a comprehensive review of the literature. Am J Geriatr Pharmacother 2011;9:405–412.
Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1. Nine gamma functions used as HRF to explore a time window starting 10 s before the PPR onset up until 10 s after that. Table S1. Demographic and clinical characteristics of the JME cohort. Table S2. Between-group comparisons of BOLD signal changes evoked by IPS: positive BOLD response. Table S3. Between-group comparisons of BOLD signal changes evoked by IPS: negative BOLD response.
Epilepsia, 55(7):1038–1047, 2014 doi: 10.1111/epi.12634
