Epilepsy Research (2014) 108, 1286—1298
journal homepage: www.elsevier.com/locate/epilepsyres
Recent seizure activity alters motor organization in frontal lobe epilepsy as revealed by task-based fMRI Kristine E. Woodward a,e,f, Ismael Gaxiola-Valdez e,f, David Mainprize f, Matthew Grossi f, Bradley G. Goodyear c,d,e,f, Paolo Federico a,b,c,e,f,∗ a
Department of Neuroscience, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 b Department of Clinical Neuroscience, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 c Department of Radiology, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 d Department of Psychiatry, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 e Hotchkiss Brain Institute, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 f Seaman Family MR Research Center, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 Received 6 February 2014; received in revised form 26 May 2014; accepted 13 June 2014 Available online 6 July 2014
KEYWORDS Connectivity; Motor coordination; BOLD; Frontal lobe epilepsy; fMRI
Summary Purpose: Patients with frontal lobe epilepsy (FLE) commonly demonstrate motor impairments, suggesting that frontal lobe seizures affect motor function. However, the underlying mechanisms of these deficits are not known, nor has any study systematically examined motor organization in these patients. We therefore examined cortical motor organization in a group of adult patients with FLE, using task-based fMRI. Methods: Eleven right FLE patients, six left FLE patients, and ten control subjects underwent task-based fMRI. Two tasks were performed using the right and left hands separately, and both hands together. The first task was a finger-tapping task and the second task was a more complex coordination task. Functional MR data were compared between patient groups and controls. A laterality index of brain activation was also calculated between the epileptic and healthy hemisphere to determine hemispheric dominance during task performance to explore its relationship with a variety of patient-specific epilepsy factors.
∗ Corresponding author at: Departments of Clinical Neurosciences and Radiology, University of Calgary, Room C1214a, Foothills Medical Centre, 1403 29th Street NW, Calgary, AB, Canada T2N 2T9. Tel.: +1 403 944 4091; fax: +1 403 283 2270. E-mail addresses: [email protected] (K.E. Woodward), [email protected] (I. Gaxiola-Valdez), [email protected] (D. Mainprize), [email protected] (M. Grossi), [email protected] (B.G. Goodyear), [email protected] (P. Federico).
http://dx.doi.org/10.1016/j.eplepsyres.2014.06.015 0920-1211/© 2014 Elsevier B.V. All rights reserved.
The effect of FLE on motor organization
1287
Results: Overall, right FLE patients demonstrated decreased BOLD activity in the epileptic hemisphere and increased BOLD activity in the healthy hemisphere compared to controls (p < 0.05). The comparison of left FLE patients to controls provided less conclusive differences, possibly due to the low number of left FLE patients studied. Laterality indices of the coordination task were positively correlated to the number of months since the last seizure in both patient groups (right FLE: rs = 0.779, left FLE: rs = 0.943). Patients that had experienced a recent seizure relied more on the sensorimotor cortex of the healthy hemisphere during task performance, compared to those that were relatively seizure free (p < 0.05). Significance: Patients with FLE exhibited changes in motor BOLD activity that was dependent on the duration of seizure freedom. These results demonstrate the presence of seizure-related alteration of cortical motor organization in FLE, which may underlie the motor deficits seen in these patients. © 2014 Elsevier B.V. All rights reserved.
Introduction Frontal lobe epilepsy (FLE) is the second most common form of focal epilepsy (Tellez-Zenteno et al., 2005). For most types of frontal lobe seizures, motor symptoms occur during the ictal period; however, patients with FLE exhibit deficits in motor control, coordination, and planning during daily activities, suggesting that seizures also affect motor function during interictal periods (Exner et al., 2002; Helmstaedter et al., 1996). This is further demonstrated by poorer performance on tests of psychomotor speed, coordination, motor planning, and motor sequencing in FLE patients when compared to patients with temporal lobe epilepsy (Helmstaedter et al., 1996; Upton and Thompson, 1996). Despite these findings, the mechanisms underlying motor deficits in FLE remain unknown. Few studies have investigated the cortical organization of motor function in FLE patients. Direct cortical stimulation and transcranial magnetic stimulation studies have revealed alterations in motor organization, but with variable results (Branco et al., 2003; Hamer et al., 2005; Labyt et al., 2007). A limitation of these studies was poor spatial resolution compared to other available imaging techniques, namely functional MRI (fMRI). Not only does fMRI provide high spatial resolution, but it is also non-invasive with no known risks, and is therefore a mainstay in the study of brain organization. Several case studies of patients with epilepsy have demonstrated changes in cortical motor organization using task-based fMRI (Macdonell et al., 1999; Moo et al., 2002; Stoeckel et al., 2002). An intracortical microstimulation study of rats also showed that repeated seizure activity near the sensorimotor cortex leads to impaired interictal motor performance, along with alterations of motor cortex maps (Henry et al., 2008). Additionally, we recently examined the effects of FLE on cortical motor networks as examined using resting-state fMRI and found that increased seizure burden was correlated with decreases in functional connectivity within the motor network (Woodward et al., 2014). These studies suggest that frontal lobe seizures can induce changes in motor cortex organization, and that these changes may be related to interictal motor deficits. The organization of cortical motor function has also been shown to change over time in epilepsy patients and animal models of epilepsy. In a case study of FLE, fMRI scans taken before and after multiple subpial transections of right
motor cortex showed increased motor activation in the unaffected hemisphere seven weeks post surgery, and a return to pre-surgical activation 13 weeks after surgery (Moo et al., 2002). A study using the pilocarpine rat model of epilepsy also demonstrated variable patterns of motor cortex organization, with larger cortical motor representations 48 h following a seizure, and a return to control patterns one and three weeks later (Young et al., 2009). These studies suggest that motor organization may be affected by individual epilepsy factors, such as recent seizures or surgical procedures. To date, no fMRI studies have examined motor organization in a group of adult patients with FLE, largely because brain activation is strongly dependent on seizure focus location, seizure frequency, age at epilepsy diagnosis, and other factors (Muller et al., 1997; Upton and Thompson, 1997). We therefore studied motor organization in FLE patients, as well as the effect of individual clinical factors on this organization. Patterns of brain activation also depend on lesion type (Janszky et al., 2003), functional deficit zone as measured by FDG-PET (Gaillard et al., 2011), and EEG findings of interictal activity (Janszky et al., 2006). These data were not available for all patients in the present study, so these factors were not investigated. We hypothesized that FLE would be associated with decreased activation in the epileptic hemisphere and increased activation in the non-epileptic hemisphere. We also hypothesized that these changes would be more pronounced in individuals with factors associated with increased seizure burden (e.g., earlier epilepsy diagnosis, longer disease duration, higher seizure frequency, etc.). Determining the relationship between motor impairments, brain motor organization, and seizure burden factors may help better understand the underlying mechanisms of motor impairments experienced in FLE patients.
Materials and methods Participants This study was approved by the Conjoint Health Research Ethics Board of our institution. Written informed consent was obtained from all participants prior to participation. Patients with FLE were identified through the Alberta Health Services — Calgary Zone Epilepsy Clinic, the Seizure
1288
Table 1 Gender
FLE patient characteristics. Age at scan
Patients with right FLE M 16
Age at seizure onset
Seizure burden
Months since last seizure
Seizure focus
Seizure types
MRI
FDG-PET hypometabolism
Neuro-psychology testing
7
Low
7
Supplementary sensorimotor
GTC, FDC
Not performed
Mild right frontal lobe dysfunction
Unremarkable
Mild right frontal lobe dysfunction
Not performed
Not performed
20
12
Moderate
0
Primary motor
GTC, FDC, FNDC
F
21
18
Moderate
2
GTC
M
24
4
Moderate
8
FNDC
Unremarkable
Not performed
Not performed
M
25
17
Low
32
Anterior frontopolar Primary motor Anterior frontopolar
GTC
Not performed
Not performed
F
29
5
Very high
0
Opercular
FDC
hypometabolism in the inferior frontal region corresponding to MRI abnormality
Mild right frontotemporal dysfunction
M
32
26
Moderate
14
Anterior frontopolar
FNDC
Not performed
Not performed
M
33
28
Low
20
Anterior frontopolar
FDC
Not performed
Not performed
F
36
0
Low
10
GTC, FDC
Not performed
Not performed
M
46
39
Low
13
Not performed
47
28
Low
112
Small vessel ischemic changes Right frontal cavernoma
Not performed
M
Supplementary sensorimotor Anterior frontopolar Supplementary sensorimotor
Small T2 signal hyperintensity in right anterior superior frontal cortex Cortical thickening right anterior insula and inferior lateral frontal love near the operculum CT shows right frontal cystic encephalomalacia Rt posterior fossa arachnoid cyst (incidental) Unremarkable
Not performed
Not performed
Mean ± SD
29.9 ± 10.2 16.7 ± 12.4
19.8
GTC, FDC GTC
K.E. Woodward et al.
F
Focal abnormality in the subcortical superior aspect of the frontal lobe Nonspecific right subcortical frontal T2 hyperintensity Unremarkable
4
Low
6
Supplementary sensorimotor Primary motor Supplementary sensorimotor Dorsolateral
GTC, FDC, FNDC GTC, FDC, FNDC GTC, FNDC
Unremarkable
Not performed
Not performed
Unremarkable
Not performed
Not performed
Unremarkable
Not performed
GTC, FDC
Unremarkable
Small area of gliosis in left primary motor cortex Encephalomalacia (post-traumatic) in left lateral frontal lobe at vertex and in left anteriolateral temporal lobe
Hypometabolism in left perisylvian frontal lobe Not performed
Mild left frontal lobe dysfunction Left frontal lobe dysfunction
M
28
2
Very high
2
M
30
8
High
0
M
39
2
Very high
1
F
55
12
Low
18
Primary motor
GTC, FNDC
F
65
41
Low
30
Primary motor
CP, FNDC
Mean ± SD
39.3 ± 17.5 11.5 ± 15.0
Left anterior temporal hypometabolism
The effect of FLE on motor organization
Patients with left FLE M 19
Not performed
Not performed
9.5
Seizure types are listed as FNDC (focal without dyscognitive features), FDC (focal with dyscognitive features), GTC (generalized tonic-clonic). Seizure burden was evaluated at the time of study as low (seizures every six months or longer), moderate (every one to six months), high (every two to four weeks), or very high (every two weeks or less).
1289
1290 Monitoring Unit of the Foothills Medical Center, and the Comprehensive Epilepsy Program, in Calgary, AB. Recruitment and scanning took place concurrently between May of 2010 and February of 2013. Twenty-three patients were recruited and scanned. Six patients were excluded due to (1) inability to complete the study (one patient), (2) being left-hand dominant, as determined by identifying the dominant writing hand (three patients), and (3) having a change in epilepsy diagnosis from FLE to temporal lobe epilepsy based on subsequent video-EEG monitoring investigations (two patients). This left 17 patients: 11 right FLE (7M, 4F; age = 29.9 ± 10.2 years) and six left FLE (4M, 2F; age = 39.3 ± 17.5 years) (Table 1). The number of patients studied was a sample of convenience based on the available patients. A diagnosis of FLE was confirmed by history, examination, routine EEG, anatomical MR imaging, and in 8 patients, scalp video-EEG monitoring. Ten healthy controls (5M, 5F; age = 33.9 ± 12.7 years) were recruited through word-of-mouth, and had no known neurological or psychiatric disorders. Other exclusion criteria included contraindication to MR imaging [including pregnancy, severe claustrophobia, metallic foreign bodies in the eye, and certain implanted medical devices (e.g., cardiac pacemaker, aneurysm clip)] and previous brain surgery. Two patients underwent surgical resection following the study and pathology showed microdysgenesis in one (left FLE, male, age 30 years) and mild gliosis in the other (right FLE, female, age 29 years). All participants were righthanded.
Behavioral motor testing Participants performed motor tasks by following a visual cue while in the MR scanner. A video projector situated behind the scanner projected visual stimuli onto a screen, which participants viewed using a mirror mounted on the head coil. Participants performed two motor tasks similar to Luria’s Motor Sequences (Luria and Haigh, 1973). The first, a fingertapping task, was used due to its simplicity and ease of use for both healthy participants and those with motor impairments. Participants tapped their fingers in time with a visual cue (contrasting colors alternating at 1-s intervals), and each run consisted of alternating blocks of rest (24 s) and task (12 s), for a total of 3 min and 48 s. The task was performed in separate runs: unimanually with the left and right hand, and bimanually with synchronous tapping. While finger-tapping elicits a brain response predominantly within primary motor regions, more complex tasks elicit strong activity within secondary and tertiary motor regions (Leonard et al., 1988). This recruitment pattern is more representative of daily functions. Therefore, a motor coordination task was also used, which required participants to perform a series of hand movements in time with a visual stimulus (circle = hand held in fist, vertical line = hand held vertically, horizontal line = hand held horizontally) paced at 2-s intervals. Each run consisted of alternating blocks of rest (24 s) and task (24 s) for a total of 5 min. Tasks were performed in separate runs: unimanually with the left and right hand, and bimanually. During the bimanual coordination task, participants performed different hand movements with each hand simultaneously.
K.E. Woodward et al. The order of the six runs was randomized for each participant. During rest periods for both tasks, participants focused on a central fixation cross. Participants practiced the tasks outside of the scanner until they were comfortable performing them. Video recordings of the coordination task were made for subsequent analysis of performance. Participants’ performance was rated blindly using a four point scale (Helmstaedter et al., 1996), ranging from one (no impairment) to four (severe impairment; Supplementary Table 1). A Mann—Whitney-U test was used to identify significant performance differences between patients (right and left FLE together) and controls. Additionally, a Kruskal—Wallis test was used to identify significant performance differences between right FLE, left FLE, and controls. Individual Mann—Whitney-U tests were then conducted between all group pairs, and a Bonferroni correction was applied to correct for multiple comparisons.
fMRI data collection Participants were asked to keep as still as possible, and their head was immobilized using foam cushioning. They could terminate the study at any time during the scan using a squeeze ball placed by their side. MR data were acquired using a 3.0 T GE Discovery MR750 whole body scanner (GE Healthcare, Waukesha, WI) with a receive-only 8-channel phased-array head coil (8 Channel High Resolution Brain Array, distributed by GE Medical Systems). MR images providing BOLD contrast were collected using a gradient-recalled echo, echo planar imaging sequence (voxel dimensions 3.75 × 3.74 × 4 mm, 28 slices, 4 mm slice thickness, 64 × 64 matrix, TE = 30 ms, TR = 1.5 s, flip angle = 65 degrees). In total, six randomized functional MRI scans were completed: one for each finger-tapping task (left, right, bimanual), and one for each coordination task (left, right, bimanual). Participants also underwent T1-weighted multi-slice spoiled gradient echo (28 × 4 mm slices, 128 × 128 matrix, minimum TE, TR = 150 ms, flip angle = 18 degrees) and 3D magnetization-prepared gradient-echo sequences (2 mm slices, 384 × 256 × 112 matrix, preparation time = 500 ms, minimum TE, TR = 8.9 ms, flip angle = 20 degrees) for anatomical registration of the fMRI data.
Preprocessing Preprocessing of image data was performed using FSL [fMRIB Software Library; http://www.fmrib.ox.ac.uk/fsl/] (Jenkinson et al., 2012). These steps included brain extraction using the Brain Extraction Tool (Smith, 2002), correction for interleaved slice timing using Fourier-space time-series phase-shifting, motion correction using MCFLIRT [Motion Correction: FMRIB’s Linear Image Registration Tool] (Jenkinson et al., 2002), spatial smoothing with a 6 mm FWHM Gaussian kernel, high pass temporal filtering (Gaussian-weighted least-squares straight line fitting with sigma = 100.0 s), grand-mean intensity normalization of the entire 4D dataset using a single multiplicative factor, and registration to the common brain template of the Montreal Neurological Institute (MNI152 standard brain, © 1993—2009
The effect of FLE on motor organization Louis Collins, McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University).
First-level analysis Models were created for each task by convolving the binary boxcar function (indicating the timing of task blocks) with the standard hemodynamic response. Models were then used in a time-series General Linear Model analysis using FEAT (Jenkinson et al., 2012) to determine brain regions that were significantly more active during task performance than rest in each participant.
Higher-level analysis Group analyses of parameter estimates were conducted to obtain mean BOLD activity maps for each task in each group. This analysis was carried out in FSL using FEAT, which utilizes the General Linear Mixed Model for higher-level analyses. This process registered each first-level analysis’ parameter estimates to the MNI152 standard brain in order to compute the average response magnitude for each voxel. Specifically, FSL’s FLAME (FMRIB’s Local Analysis of Mixed Effects) was used, which takes the session and participant variability into account, therefore allowing inferences to be made about the wider population and not only about individuals whom participated in the study. Contrasts of parameter estimates were generated to compute mean BOLD activity for each group and between groups and tasks. Specifically, the General Linear Mixed Model was applied by FSL to determine voxels exhibiting a significant difference in response magnitude between groups and tasks. Participant’s coordination performance scores were demeaned and entered as a covariate when computing both the mean and contrast images involving the coordination data. Maps were generated using a corrected cluster significance of p = 0.05, using AlphaSim, which uses Monte Carlo simulations of null distribution image data to estimate family-wise error rates (Ward, 2000).
Laterality indices A laterality index (LI) was used as an indicator of hemispheric dominance (in terms of BOLD activity levels) during task performance. LI was calculated for each individual’s left-handed and right-handed coordination and fingertapping maps, as follows: LI =
1291 they could occur in any cortical region and not just regions commonly thought of as ‘cortical motor areas’. Because the whole-hemisphere was used in the calculations, solely comparing the number of voxels would not quantify changes in BOLD activity levels within each hemisphere, and would just be a measure of brain size. Thus, average Z score was used in the calculation to determine hemispheric dominance during task performance. Control participants LIs were calculated in the same manner as the FLE group that it was being compared to. For example, when controls and right FLE patients were compared using the right-handed coordination task, ‘epileptic’ was the right hemisphere in both groups, and ‘healthy’ was the left hemisphere in both groups. A positive LI value indicated epileptic hemispheric dominance during task performance, whereas a negative LI value indicated healthy hemispheric dominance during task performance. Spearman’s rank correlation analyses were performed between LI and the following factors for each patient group: age at epilepsy diagnosis, years since diagnosis, lifetime seizures (total or generalized tonic-clonic seizures alone), seizures in past year (total or generalized tonic-clonic seizures alone), number of months since the last seizure, and coordination performance score (i.e., one, two, three, or four). Seizure burden factors were determined by retrospectively reviewing the patients’ charts.
Results Behavioral motor testing All participants received a score of one on the unimanual coordination task. On the bimanual coordination task, FLE patients performed significantly worse than control subjects (p = 0.011), in particular the left FLE patients (p = 0.024) (Table 2). The scores for each group were as follows: right FLE: 1.4 ± 0.5; left FLE 2.0 ± 0.9; control 1.0 ± 0.0. There was no significant relationship between motor performance score and length of seizure freedom.
Brain activation During all tasks (right-handed, left-handed, bimanual in both coordination and finger-tapping), control participants showed BOLD activity in cortical regions including
(# voxels epileptic)(avg Z epileptic) − (# voxels healthy)(avg Z healthy) (# voxels epileptic)(avg Z epileptic) + (# voxels healthy)(avg Z healthy)
where ‘epileptic’ was the hemisphere ipsilateral to the seizure focus, and ‘healthy’ was the hemisphere contralateral to the seizure focus. ‘# voxels’ included all voxels in the gray matter of the respective hemisphere, and ‘avg Z’ was the average Z score of all of the voxels in the gray matter of the respective hemisphere. Whole-hemisphere calculations were used because it was hypothesized that if changes in cortical organization were to occur due to epileptic activity,
the primary motor, supplementary motor, prefrontal, posterior parietal, and lateral occipital cortex (Supplementary Figure 1). Regions of BOLD activity in subcortical structures included the putamen, globus pallidus, thalamus, and caudate. Areas of reduced BOLD activity during task performance included the medial prefrontal cortex, posterior
1292
K.E. Woodward et al.
Table 2 Comparison of bimanual coordination performance scores between each of the three groups, analyzed using Kruskal—Wallis (KW) and Mann—Whitney-U (MW) tests. Significant differences (p < 0.05) are shown in bold and marked by an asterisk (*). Comparison
P-value
Bonferroni corrected p-value (when necessary)
FLE vs Controls (MW) All three groups (KW) Right FLE vs Controls (MW) Left FLE vs Controls (MW) Right FLE vs Left FLE (MW)
0.01 (1-tailed) 0.01 0.06 (1-tailed) 0.008 (1-tailed) 0.08 (2-tailed)
0.01* 0.01* 0.17 0.02* 0.45
cingulate, angular gyrus, and posterior insula, all regions that are part of the default mode network. Fig. 1 shows brain regions of right FLE patients where BOLD activity was significantly different than controls during task performance (summarized in Supplementary Table 2). During the left-handed finger-tapping task, right FLE patients showed significantly less BOLD activity in the right middle frontal gyrus, right inferior parietal cortex, and right occipital pole. During the right-handed finger-tapping task, right FLE patients had less BOLD activity in the right and left lateral occipital cortex, right lingual gyrus, and right inferior parietal cortex. During the bimanual finger-tapping task, right FLE patients had less BOLD activity in the right and left frontal pole, right insula, right lateral occipital cortex, right inferior parietal cortex, right superior parietal cortex, right putamen, right thalamus, and precuneus. During the left-handed coordination task, right FLE patients exhibited greater BOLD activity in the left inferior parietal cortex and left precentral gyrus (Fig. 1, Supplementary Table 2). During the right-handed coordination task right FLE patients also had decreased BOLD in the right lateral occipital cortex, precuneus, right superior parietal cortex, and right precentral gyrus. During the bimanual coordination task, these patients showed greater BOLD activity in the left superior and middle temporal gyri, left insular cortex, anterior cingulate, left superior parietal cortex, left superior frontal gyrus, as well as less BOLD activity in the right frontal operculum and right precentral gyrus. Fig. 2 shows brain regions of left FLE patients where BOLD activity was significantly different than controls during task performance (summarized in Supplementary Table 3). During the left-handed finger-tapping task, patients had decreased BOLD activity in the right middle frontal gyrus, right frontal pole, right lateral occipital cortex, and left frontal pole. During the right-handed finger-tapping task, patients had decreased BOLD activity in the left superior parietal lobule. During the bimanual finger-tapping task, patients had decreased BOLD activity in the right lateral occipital cortex, right superior frontal gyrus, right orbitofrontal cortex, right putamen, left frontal pole, left superior parietal lobule, and left precuneus. During the left-handed coordination task, patients had decreased BOLD activity in the right inferior parietal cortex, right middle frontal gyrus, and right putamen (Fig. 2, Supplementary Table 3). During the right-handed coordination task, patients had decreased BOLD activity in the right caudate and left insula. During the bimanual coordination
task, patients had decreased BOLD activity in the right inferior parietal cortex, and increased BOLD activity in the left superior temporal gyrus and left insula.
Relationship between LI and seizure demographics During the right-handed coordination task, control participants had an average LI of −0.02 ± 0.59 (range = −0.90 to 0.94), and an average LI of 0.04 ± 0.40 (range = −0.71 to 0.60) during the left-handed coordination task. During the right-handed finger-tapping task, control participants had an average LI of −0.22 ± 0.39 (range = −0.99 to 0.54), and an average LI of −0.10 ± 0.20 (range = −0.44 to 0.11) during the left-handed finger-tapping task. None of the tasks differed significantly from 0. Patients with right FLE had an average LI of 0.97 ± 1.59 (range = −1.68 to 3.98) during the left-handed coordination task, with five patients falling either above or below the control (‘normal’) LI range, and an average LI of 0.22 ± 0.55 (range = −0.55 to 0.99) during the left-handed finger-tapping task. Patients with left FLE had an average LI of 0.10 ± 0.30 (range = −0.40 to 0.41) during the righthanded coordination task, and an average LI of 0.55 ± 0.43 (range = 0.20 to 0.92) during the right-handed finger-tapping task. There was no significant difference between LIs of right or left FLE patients and controls. All participants received a performance score of one on the unimanual coordination tasks. Therefore, when determining the effects of motor impairments on LI, only the bimanual coordination performances scores were compared. Patients with a score of one (n = 9), two (n = 6), and three (n = 3) had an average LI of 0.99 ± 1.33, 0.48 ± 1.52, and −0.23 ± 0.24, respectively. No patients received a score of 4. There was no significant difference between groups. In FLE patients, a positive correlation was seen between the number of months since the last seizure, and the LI of the coordination task that employed the epileptic hemisphere (i.e., left-handed coordination task for right FLE patients, rS = 0.779; right-handed coordination task for left FLE patients, rS = 0.943; Fig. 3). No correlation was observed between LI and the coordination task that would typically employ the healthy hemisphere, or between LI and the finger-tapping tasks. Fig. 3 also shows the relationship between the duration of seizure freedom and brain activation during the unimanual coordination task. Specifically, with longer periods of seizure freedom, right FLE patients showed less BOLD signal in the left sensorimotor cortex, and
The effect of FLE on motor organization
1293
Figure 1 Comparison of brain activation between right FLE patients and controls during finger-tapping and coordination tasks. Areas of red/yellow indicate regions that were more active in right FLE patients compared to controls during task performance. Areas of blue/light blue indicate regions that were less active in right FLE patients compared to controls during task performance.
left and right inferior parietal cortex, and left FLE patients showed less BOLD signal in the right sensorimotor cortex, and left and right occipital cortex (summarized in Supplementary Table 4).
Discussion This was the first fMRI study to compare motor activation between a group of adult FLE patients and healthy controls.
Right FLE patients had significant decreases in activation of the epileptic hemisphere, and significant increases in activation of the healthy hemisphere during task performance. Results were less conclusive in the left FLE patients, possibly due to a limited number of patients. Additionally, in both patient groups there was a positive correlation between sensorimotor cortex activation in the epileptic hemisphere and the duration of seizure freedom prior to the study.
1294
K.E. Woodward et al.
Figure 2 Comparison of brain activation between left FLE patients and controls during finger-tapping and coordination tasks. Areas of red/yellow indicate regions that were more active in left FLE patients compared to controls during task performance. Areas of blue/light blue indicate regions that were less active in left FLE patients compared to controls during task performance.
Finger-tapping tasks During unimanual finger-tapping tasks, both patients groups showed decreased BOLD activity compared to controls (Figs. 1 and 2). One region common to both group comparisons was the posterior parietal cortex of the epileptic hemisphere, solely when performing tasks that primarily taxed the epileptic hemisphere (i.e., left-handed task for
right FLE patients, right-handed task for left FLE patients). This may have been due to detrimental effects of seizure propagation from the ipsilateral seizure focus, causing patients to rely less on the epileptic hemisphere. Indeed, seizure propagation paths form strong neuronal networks due to synchronous and repetitive activity, which could disrupt normal networks in the brain, such as the motor network. The posterior parietal cortex is part of the motor
The effect of FLE on motor organization
1295
Figure 3 Left: relationship between LIs of the left-handed coordination task and months since last seizure in right FLE patients (rs = 0.779), and LIs of the right-handed coordination task and months since last seizure in left FLE patients (rs = 0.943). Increasingly positive LI values indicate greater dominance of the epileptic hemisphere during task performance. Right: relationship between duration of seizure freedom and brain activation during the unimanual coordination task. Images indicate brain regions that were more active (red) or less active (blue) as a function of the duration of seizure freedom for the left-handed coordination task in right FLE patients (upper row) and the right-handed coordination task in left FLE patients (lower row).
network, and is pivotal for synchronizing hand movements with external stimuli, as well as integrating sensory and motor information (Iacoboni, 2006). While we did not observe motor impairments during unimanual finger-tapping tasks, decreased BOLD activity in the posterior parietal cortex may in part be responsible for daily motor impairments seen in FLE patients. Similar to unimanual finger-tapping tasks, the bimanual finger-tapping task was associated with decreased BOLD activity in the posterior parietal cortex of the epileptic hemisphere in both patient groups (Figs. 1 and 2). Additionally, during the bimanual tapping task both patient groups had less BOLD activity in brain regions that are part of the default mode network. The default mode network is primarily activated in the absence of external stimuli, and brain activity is increasingly suppressed during more complex tasks (McKiernan et al., 2003). This suggests that patients focused more on external stimuli than did controls, possibly to overcome motor impairments in order to perform the task fluently.
Coordination tasks During unimanual coordination tasks, right FLE patients had increased BOLD signal in motor regions of the healthy hemisphere during the task that taxed the epileptic hemisphere (left-handed coordination, Fig. 1). This suggests that normal recruitment of the epileptic hemisphere was insufficient to perform the left-handed task, and that additional, healthy hemisphere brain regions were recruited as a means of compensation. However, this compensatory activation was not as effective as normal activation patterns because the patient group did not perform as well as the control group (Table 2). This suggests that while healthy hemisphere brain regions
were necessary for patients to complete the task, they were not as effective in eliciting correct hand maneuvers as the cortex that was originally responsible for such movements. Furthermore, right FLE patients had decreased BOLD signal in motor regions of the epileptic hemisphere during the task that primarily taxed the healthy hemisphere (righthanded coordination). This suggests that the right-handed task rendered compensatory activity unnecessary for adequate motor performance, as the epileptic hemisphere was not being recruited as heavily. The comparison of unimanual coordination tasks between left FLE and controls produced more inconsistent results with no clear trends. We suspect that this variability may be due to the limited number of left FLE patients included in the study. During the bimanual coordination task, both patient groups had regions of increased left hemisphere BOLD activity (Figs. 1 and 2). This suggests that these changes in activation might not be related to seizure focus lateralization, but rather hemispheric dominance. All participants were right-handed; therefore, these findings suggest increased reliance on regions in the left, or dominant, hemisphere when compared to controls. Dominant hemisphere connections are generally stronger because of increased use (Chen et al., 1997; Kim et al., 1993); therefore, they may be less susceptible to detrimental seizure activity. If non-dominant hemisphere connections were impacted by seizure activity, this may cause patients to rely more on the dominant hemisphere during motor task performance, as was seen in the bimanual coordination task. During bimanual coordination, increased BOLD signal was seen in the left insula and superior temporal gyrus in both FLE groups (Figs. 1 and 2). Right FLE patients also had increased BOLD activity in the left anterior cingulate. All of these regions have been implicated in error-detection, and show increased BOLD signal during complex motor
1296 tasks when compared to simpler tasks (Carter et al., 1998; Debaere et al., 2004; Ullsperger and von Cramon, 2003; Wendelken et al., 2009). This suggests that patients made more errors, presumably due to the difficulty of the bimanual coordination task. Indeed, patients had significantly poorer performance scores on the bimanual coordination task when compared to controls (Table 2). Additionally, increased BOLD activity was observed in the superior temporal gyrus, which is activated during sequence processing tasks (Ullen et al., 2005), such as the coordination task in this study. However, less activation was seen with increased practice (Jantzen et al., 2002), again suggesting that patients had not mastered the task as well as control participants.
Relationship between LI and period of seizure freedom During coordination tasks that taxed the epileptic hemisphere, patients with recent seizures had increased BOLD activity in the sensorimotor cortex of the healthy hemisphere, whereas patients with longer periods of seizure freedom had less BOLD activity in this region (Fig. 3). This suggests that propagation of abnormal neuronal activity, due to recent seizures, adversely affected the epileptic hemisphere, thereby causing patients to rely more on sensorimotor cortex of the healthy hemisphere. These findings are supported by studies of motor recovery following stroke and epilepsy surgery (multiple subpial transection of the seizure focus). Patients with strokes or seizure foci in primary motor cortex initially had increased BOLD activity in the healthy hemisphere when using the affected hand, which persisted over many weeks; however, there was an eventual return to pre-stroke, or pre-surgical activation patterns (Chollet et al., 1991; Moo et al., 2002; Feydy et al., 2002). Furthermore, patients demonstrated no changes in brain activation over time when testing the unaffected hand (Chollet et al., 1991; Moo et al., 2002). This is similar to the results of our study, in which no correlation was seen between LI of the task using the unaffected hand and period of seizure freedom.
Pathophysiology of changes in motor cortex organization Given the relatively rapid changes in brain activation described in the previous section, the reversibility of these changes, and distance of changes from the lesion site, it is likely that pre-existing but dormant connections were in place to facilitate these changes (Feydy et al., 2002; Moo et al., 2002). Indeed, horizontal cortico-cortical connections are known to exist between primary motor cortex and other motor regions, both within and between hemispheres (Castro-Alamancos and Borrel, 1995; Feydy et al., 2002). If such connections became less active following a lesion or recurrent seizures, this could result in disinhibition, thereby allowing recruitment of additional sites to maintain normal motor function. Indeed, patients with more frequent seizures have impaired cortico-cortical inhibition as measured by transcranial magnetic stimulation (Labyt et al., 2007). However, once normal conditions are reinstated, or
K.E. Woodward et al. perhaps long periods of seizure freedom ensue, inhibition can be re-exerted upon compensatory cortical regions and normal activation patterns return (Feydy et al., 2002; Moo et al., 2002). Thus, recent seizures could initially increase recruitment of healthy hemisphere regions during coordination tasks, which would return back to baseline following longer periods of seizure freedom (Fig. 3). Previous studies using rat models of epilepsy have also suggested that disinhibition may underlie seizure-induced changes in cortical motor map representations (CastroAlamancos and Borrel, 1995; van Rooyen et al., 2006; Young et al., 2009). Both decreases in GABAergic and increases in glutamatergic transmission, which occur in humans with epilepsy, have been shown to influence the recruitment of additional motor regions through pre-existing corticocortical connections (Behr et al., 2000; Hess et al., 1996; van Rooyen et al., 2006). Additionally, in rats, repeated kindling near the sensorimotor cortex leads to impaired interictal behavioral motor performance (Henry et al., 2008). Such altered motor performance was associated with more efficacious synapses, providing further evidence for disinhibition as an underlying mechanism of impaired motor function. This is shown in our results as patients with the worst performance scores had the lowest average LI values, indicating greater reliance (possibly through disinhibition) on the healthy hemisphere. As performance scores improved, LI values increased to indicate greater reliance on the epileptic hemisphere, conceivably by establishing a more normal pattern of BOLD activity through a re-exertion of inhibition on healthy hemisphere connections.
Limitations of study We did not study a large number of participants, due to the limited number of FLE patients who were candidates for the study. Increasing the number of participants might further clarify our results, as we would be able to separate patients into FLE patient subgroups based on the specific location of the seizure focus within the frontal lobes, amongst other patient-specific factors. Indeed, the location of the seizure foci within the frontal lobe (Koudijs et al., 2010) as well as seizure frequency, severity, and duration (Upton and Thompson, 1996) affect the results of task-based fMRI. However, we attempted to account for patient heterogeneity by determining the effect of a variety of patient-specific factors on LIs. When we did this, we found a positive correlation between the number of months since the last seizure and the LI for the coordination task in both patient groups. Despite the average age difference between groups, there was no statistically significant difference in age. A previous study that demonstrated differences in motor activity due to age examined two groups of patients with average ages of 23 and 66 for the younger and older groups, respectively, a difference of 43 years (Riecker et al., 2006). This is much larger then the maximum difference of 10 years in our study. There were also small differences in the number of males and female in each group; however, other studies of gender differences in motor activation during task performance have shown that there is no difference in activation patterns between males and females (Ward and Frackowiak, 2003).
The effect of FLE on motor organization Patients were taking a variety of anticonvulsant medications in combination or isolation. Previous studies have shown that these drugs can decrease BOLD activity during task-based fMRI studies (Jansen et al., 2006; Szaflarski and Allendorfer, 2012), or have no significant effect on BOLD activity (Braakman et al., 2013; Yasuda et al., 2013). We also confirmed that anticonvulsant medications have no significant effect on the BOLD signal during task-based motor fMRI (unpublished observations), which suggests that the results of the present study were not likely related to medication effect.
Conclusions This was the first known study to examine motor fMRI activation in a group of adult patients with FLE. In general, patients showed significantly greater BOLD activity in brain regions contralateral to the seizure focus, and significantly less BOLD activity in brain regions ipsilateral to the seizure focus while performing a motor task. Additionally, patients with recent seizures had an initial reliance on the healthy hemisphere, and an eventual recovery of epileptic hemisphere BOLD activity with longer periods of seizure freedom. These results suggest that changes in cortical activation may be an underlying mechanism of motor deficits observed in FLE patients, and suggest the potential for longer periods of seizure freedom to reduce these motor deficits.
Conflict of interest None of the authors have 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.
Acknowledgements This work was supported by CIHR (MOP-230809). We also thank Daniel J Pittman and Aaron Spring for critically reviewing the manuscript.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.eplepsyres.2014.06.015.
References Behr, J., Heinemann, U., Mody, I., 2000. Glutamate receptor activation in the kindled dentate gyrus. Epilepsia 41 (Suppl. 6), S100—S103. Braakman, H.M., Vaessen, M.J., Jansen, J.F., Debeij-van Hall, M.H., de Louw, A., Hofman, P.A., Vles, J.S., Aldenkamp, A.P., Backes, W.H., 2013. Frontal lobe connectivity and cognitive impairment in pediatric frontal lobe epilepsy. Epilepsia 54, 446—454. Branco, D.M., Coelho, T.M., Branco, B.M., Schmidt, L., Calcagnotto, M.E., Portuguez, M., Neto, E.P., Paglioli, E., Palmini, A., Lima, J.V., Da Costa, J.C., 2003. Functional variability of the
1297 human cortical motor map: electrical stimulation findings in perirolandic epilepsy surgery. J. Clin. Neurophysiol. 20, 17—25. Carter, C.S., Braver, T.S., Barch, D.M., Botvinick, M.M., Noll, D., Cohen, J.D., 1998. Anterior cingulate cortex, error detection, and the online monitoring of performance. Science 280, 747—749. Castro-Alamancos, M.A., Borrel, J., 1995. Functional recovery of forelimb response capacity after forelimb primary motor cortex damage in the rat is due to the reorganization of adjacent areas of cortex. Neuroscience 68, 793—805. Chen, R., Gerloff, C., Hallett, M., Cohen, L.G., 1997. Involvement of the ipsilateral motor cortex in finger movements of different complexities. Ann. Neurol. 41, 247—254. Chollet, F., DiPiero, V., Wise, R.J., Brooks, D.J., Dolan, R.J., Frackowiak, R.S., 1991. The functional anatomy of motor recovery after stroke in humans: a study with positron emission tomography. Ann. Neurol. 29, 63—71. Debaere, F., Wenderoth, N., Sunaert, S., Van Hecke, P., Swinnen, S.P., 2004. Changes in brain activation during the acquisition of a new bimanual coodination task. Neuropsychologia 42, 855—867. Exner, C., Boucsein, K., Lange, C., Winter, H., Weniger, G., Steinhoff, B.J., Irle, E., 2002. Neuropsychological performance in frontal lobe epilepsy. Seizure: J. Br. Epilepsy Assoc. 11, 20—32. Feydy, A., Carlier, R., Roby-Brami, A., Bussel, B., Cazalis, F., Pierot, L., Burnod, Y., Maier, M.A., 2002. Longitudinal study of motor recovery after stroke: recruitment and focusing of brain activation. Stroke: J. Cereb. Circ. 33, 1610—1617. Gaillard, W.D., Berl, M.M., Duke, E.S., Ritzl, E., Miranda, S., Liew, C., Finegersh, A., Martinez, A., Dustin, I., Sato, S., Theodore, W.H., 2011. fMRI language dominance and FDG-PET hypometabolism. Neurology 76, 1322—1329. Hamer, H.M., Reis, J., Mueller, H.H., Knake, S., Overhof, M., Oertel, W.H., Rosenow, F., 2005. Motor cortex excitability in focal epilepsies not including the primary motor area — a TMS study. Brain: J. Neurol. 128, 811—818. Helmstaedter, C., Kemper, B., Elger, C.E., 1996. Neuropsychological aspects of frontal lobe epilepsy. Neuropsychologia 34, 399—406. Henry, L.C., Goertzen, C.D., Lee, A., Teskey, G.C., 2008. Repeated seizures lead to altered skilled behaviour and are associated with more highly efficacious excitatory synapses. Eur. J. Neurosci. 27, 2165—2176. Hess, G., Aizenman, C.D., Donoghue, J.P., 1996. Conditions for the induction of long-term potentiation in layer II/III horizontal connections of the rat motor cortex. J. Neurophysiol. 75, 1765—1778. Iacoboni, M., 2006. Visuo-motor integration and control in the human posterior parietal cortex: evidence from TMS and fMRI. Neuropsychologia 44, 2691—2699. Jansen, J.F., Aldenkamp, A.P., Marian Majoie, H.J., Reijs, R.P., de Krom, M.C., Hofman, P.A., Eline Kooi, M., Nicolay, K., Backes, W.H., 2006. Functional MRI reveals declined prefrontal cortex activation in patients with epilepsy on topiramate therapy. Epilepsy Behav. 9, 181—185. Jantzen, K.J., Steinberg, F.L., Kelso, J.A., 2002. Practicedependent modulation of neural activity during human sensorimotor coordination: a functional magnetic resonance imaging study. Neurosci. Lett. 332, 205—209. Janszky, J., Ebner, A., Kruse, B., Mertends, M., Jokeit, H., Seitz, R.J., Witte, O.W., Tuxhorn, I., Woermann, F.G., 2003. Functional organization of the brain with malformations of cortical development. Ann. Neurol. 53, 759—767. Janszky, J., Mertens, M., Janszky, I., Ebner, A., Woermann, F.G., 2006. Left-sided interictal epileptic activity induces shift of language lateralization in temporal lobe epilepsy: an fMRI study. Epilepsia 47, 921—927. Jenkinson, M., Bannister, P., Brady, M., Smith, S., 2002. Improved optimization for the robust and accurate linear registration and motion correction of brain images. NeuroImage 17, 825—841.
1298 Jenkinson, M., Beckmann, C.F., Behrens, T.E., Woolrich, M.W., Smith, S.M., 2012. FSL. NeuroImage 62, 782—790. Kim, S.G., Ashe, J., Hendrich, K., Ellermann, J.M., Merkle, H., Ugurbil, K., Georgopoulos, A.P., 1993. Functional magnetic resonance imaging of motor cortex: hemispheric asymmetry and handedness. Science 261, 615—617. Koudijs, S.M., Leijten, F.S., Ramsey, N.F., van Nieuwenhuizen, O., Braun, K.P., 2010. Lateralization of motor innervation in children with intractable focal epilepsy — a TMS and fMRI study. Epilepsy Res. 90, 140—150. Labyt, E., Houdayer, E., Cassim, F., Bourriez, J.L., Derambure, P., Devanne, H., 2007. Motor representation areas in epileptic patients with focal motor seizures: a TMS study. Epilepsy Res. 75, 197—205. Leonard, G., Milner, B., Jones, L., 1988. Performance on unimanual and bimanual tapping tasks by patients with lesions of the frontal or temporal lobe. Neuropsychologia 26, 79—91. Luria, A.R., Haigh, B., 1973. The working brain. (London) Allen Lane the Penguin Press, Penguin. Macdonell, R.A., Jackson, G.D., Curatolo, J.M., Abbott, D.F., Berkovic, S.F., Carey, L.M., Syngeniotin, A., Fabinyi, G.C., Scheffer, I.E., 1999. Motor cortex localization using functional MRI and transcranial magnetic stimulation. Neurology 53, 1462—1467. McKiernan, K.A., Kaufman, J.N., Kucera-Thompson, J., Binder, J.R., 2003. A parametric manipulation of factors affecting task-induced deactivation in functional neuroimaging. J. Cogn. Neurosci. 15, 394—408. Moo, L.R., Slotnick, S.D., Krauss, G., Hart, J., 2002. A prospective study of motor recovery following multiple subpial transections. Neuroreport 13, 665—669. Muller, R.A., Rothermel, R.D., Behen, M.E., Muzik, O., Chakraborty, P.K., Chugani, H.T., 1997. Plasticity of motor organization in children and adults. Neuroreport 8, 3103—3108. Riecker, A., Groschel, K., Ackermann, H., Steinbrink, C., Witte, O., Kastrup, A., 2006. Functional significance of age-related differences in motor activation patterns. NeuroImage 32, 1345—1354. Smith, S.M., 2002. Fast robust automated brain extraction. Hum. Brain Mapp. 17, 143—155. Stoeckel, M.C., Kleinschmidt, A., Ebner, A., Witte, O.W., Seitz, R.J., 2002. Reorganization of motor representation in a patient with epilepsia partialis continua as shown by [O15]-labeled butanol positron emission tomography and functional magnetic resonance imaging. J. Neuroimaging: Off. J. Am. Soc. Neuroimag. 12, 276—281.
K.E. Woodward et al. Szaflarski, J.P., Allendorfer, J.B., 2012. Topiramate and its effect on fMRI of language in patients with right or left temporal lobe epilepsy. Epilepsy Behav. 24, 74—80. Tellez-Zenteno, J.F., Dhar, R., Wiebe, S., 2005. Long-term seizure outcomes following epilepsy surgery: a systematic review and meta-analysis. Brain: J. Neurol. 128, 1188—1198. Ullen, F., Bengtsson, S.L., Ehrsson, H.H., Forssberg, H., 2005. Neural control of rhythmic sequences. Ann. N Y Acad. Sci. 1060, 368—376. Ullsperger, M., von Cramon, D.Y., 2003. Error monitoring using external feedback: specific roles of the habenular complex, the reward system, and the cingulate motor area revealed by functional magnetic resonance imaging. J. Neurosci.: Off. J. Soc. Neurosci. 23, 4308—4314. Upton, D., Thompson, P.J., 1997. Age at onset and neuropsychological function in frontal lobe epilepsy. Epilepsia 38, 1103—1113. Upton, D., Thompson, P.J., 1996. General neuropsychological characteristics of frontal lobe epilepsy. Epilepsy Res. 23, 169—177. van Rooyen, F., Young, N.A., Larson, S.E., Teskey, G.C., 2006. Hippocampal kindling leads to motor map expansion. Epilepsia 47, 1383—1391. Ward, B.D., 2000. AlphaSim: estimate statistical significance via Monte Carlo simulation. Biophysics Research Institute, Medical College of Wisconsin. Ward, N.S., Frackowiak, R.S.J., 2003. Age-related changes in the neural correlates of motor performance. Brain 126, 873—888. Wendelken, C., Ditterich, J., Bunge, S.A., Carter, C.S., 2009. Stimulus and response conflict processing during perceptual decision making. Cogn. Affect. Behav. Neurosci. 9, 434—447. Woodward, K.E., Gaxiola-Valdez, I., Goodyear, B.G., Federico, P., 2014. Frontal lobe epilepsy alters funtional connections within the brain’s motor network: a resting-state fMRI study. Brain Connect. 4, 91—99. Yasuda, C.L., Centeno, M., Vollmar, C., Stretton, J., Symms, M., Cendes, F., Mehta, M.A., Thompson, P., Duncan, J.S., Koepp, M.J., 2013. The effect of topiramate on cognitive fMRI. Epilepsy Res.. Young, N.A., Vuong, J., Ozen, L.J., Flynn, C., Teskey, G.C., 2009. Motor map expansion in the pilocarpine model of temporal lobe epilepsy is dependent on seizure severity and rat strain. Exp. Neurol. 217, 421—428.
journal homepage: www.elsevier.com/locate/epilepsyres
Recent seizure activity alters motor organization in frontal lobe epilepsy as revealed by task-based fMRI Kristine E. Woodward a,e,f, Ismael Gaxiola-Valdez e,f, David Mainprize f, Matthew Grossi f, Bradley G. Goodyear c,d,e,f, Paolo Federico a,b,c,e,f,∗ a
Department of Neuroscience, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 b Department of Clinical Neuroscience, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 c Department of Radiology, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 d Department of Psychiatry, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 e Hotchkiss Brain Institute, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 f Seaman Family MR Research Center, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4 Received 6 February 2014; received in revised form 26 May 2014; accepted 13 June 2014 Available online 6 July 2014
KEYWORDS Connectivity; Motor coordination; BOLD; Frontal lobe epilepsy; fMRI
Summary Purpose: Patients with frontal lobe epilepsy (FLE) commonly demonstrate motor impairments, suggesting that frontal lobe seizures affect motor function. However, the underlying mechanisms of these deficits are not known, nor has any study systematically examined motor organization in these patients. We therefore examined cortical motor organization in a group of adult patients with FLE, using task-based fMRI. Methods: Eleven right FLE patients, six left FLE patients, and ten control subjects underwent task-based fMRI. Two tasks were performed using the right and left hands separately, and both hands together. The first task was a finger-tapping task and the second task was a more complex coordination task. Functional MR data were compared between patient groups and controls. A laterality index of brain activation was also calculated between the epileptic and healthy hemisphere to determine hemispheric dominance during task performance to explore its relationship with a variety of patient-specific epilepsy factors.
∗ Corresponding author at: Departments of Clinical Neurosciences and Radiology, University of Calgary, Room C1214a, Foothills Medical Centre, 1403 29th Street NW, Calgary, AB, Canada T2N 2T9. Tel.: +1 403 944 4091; fax: +1 403 283 2270. E-mail addresses: [email protected] (K.E. Woodward), [email protected] (I. Gaxiola-Valdez), [email protected] (D. Mainprize), [email protected] (M. Grossi), [email protected] (B.G. Goodyear), [email protected] (P. Federico).
http://dx.doi.org/10.1016/j.eplepsyres.2014.06.015 0920-1211/© 2014 Elsevier B.V. All rights reserved.
The effect of FLE on motor organization
1287
Results: Overall, right FLE patients demonstrated decreased BOLD activity in the epileptic hemisphere and increased BOLD activity in the healthy hemisphere compared to controls (p < 0.05). The comparison of left FLE patients to controls provided less conclusive differences, possibly due to the low number of left FLE patients studied. Laterality indices of the coordination task were positively correlated to the number of months since the last seizure in both patient groups (right FLE: rs = 0.779, left FLE: rs = 0.943). Patients that had experienced a recent seizure relied more on the sensorimotor cortex of the healthy hemisphere during task performance, compared to those that were relatively seizure free (p < 0.05). Significance: Patients with FLE exhibited changes in motor BOLD activity that was dependent on the duration of seizure freedom. These results demonstrate the presence of seizure-related alteration of cortical motor organization in FLE, which may underlie the motor deficits seen in these patients. © 2014 Elsevier B.V. All rights reserved.
Introduction Frontal lobe epilepsy (FLE) is the second most common form of focal epilepsy (Tellez-Zenteno et al., 2005). For most types of frontal lobe seizures, motor symptoms occur during the ictal period; however, patients with FLE exhibit deficits in motor control, coordination, and planning during daily activities, suggesting that seizures also affect motor function during interictal periods (Exner et al., 2002; Helmstaedter et al., 1996). This is further demonstrated by poorer performance on tests of psychomotor speed, coordination, motor planning, and motor sequencing in FLE patients when compared to patients with temporal lobe epilepsy (Helmstaedter et al., 1996; Upton and Thompson, 1996). Despite these findings, the mechanisms underlying motor deficits in FLE remain unknown. Few studies have investigated the cortical organization of motor function in FLE patients. Direct cortical stimulation and transcranial magnetic stimulation studies have revealed alterations in motor organization, but with variable results (Branco et al., 2003; Hamer et al., 2005; Labyt et al., 2007). A limitation of these studies was poor spatial resolution compared to other available imaging techniques, namely functional MRI (fMRI). Not only does fMRI provide high spatial resolution, but it is also non-invasive with no known risks, and is therefore a mainstay in the study of brain organization. Several case studies of patients with epilepsy have demonstrated changes in cortical motor organization using task-based fMRI (Macdonell et al., 1999; Moo et al., 2002; Stoeckel et al., 2002). An intracortical microstimulation study of rats also showed that repeated seizure activity near the sensorimotor cortex leads to impaired interictal motor performance, along with alterations of motor cortex maps (Henry et al., 2008). Additionally, we recently examined the effects of FLE on cortical motor networks as examined using resting-state fMRI and found that increased seizure burden was correlated with decreases in functional connectivity within the motor network (Woodward et al., 2014). These studies suggest that frontal lobe seizures can induce changes in motor cortex organization, and that these changes may be related to interictal motor deficits. The organization of cortical motor function has also been shown to change over time in epilepsy patients and animal models of epilepsy. In a case study of FLE, fMRI scans taken before and after multiple subpial transections of right
motor cortex showed increased motor activation in the unaffected hemisphere seven weeks post surgery, and a return to pre-surgical activation 13 weeks after surgery (Moo et al., 2002). A study using the pilocarpine rat model of epilepsy also demonstrated variable patterns of motor cortex organization, with larger cortical motor representations 48 h following a seizure, and a return to control patterns one and three weeks later (Young et al., 2009). These studies suggest that motor organization may be affected by individual epilepsy factors, such as recent seizures or surgical procedures. To date, no fMRI studies have examined motor organization in a group of adult patients with FLE, largely because brain activation is strongly dependent on seizure focus location, seizure frequency, age at epilepsy diagnosis, and other factors (Muller et al., 1997; Upton and Thompson, 1997). We therefore studied motor organization in FLE patients, as well as the effect of individual clinical factors on this organization. Patterns of brain activation also depend on lesion type (Janszky et al., 2003), functional deficit zone as measured by FDG-PET (Gaillard et al., 2011), and EEG findings of interictal activity (Janszky et al., 2006). These data were not available for all patients in the present study, so these factors were not investigated. We hypothesized that FLE would be associated with decreased activation in the epileptic hemisphere and increased activation in the non-epileptic hemisphere. We also hypothesized that these changes would be more pronounced in individuals with factors associated with increased seizure burden (e.g., earlier epilepsy diagnosis, longer disease duration, higher seizure frequency, etc.). Determining the relationship between motor impairments, brain motor organization, and seizure burden factors may help better understand the underlying mechanisms of motor impairments experienced in FLE patients.
Materials and methods Participants This study was approved by the Conjoint Health Research Ethics Board of our institution. Written informed consent was obtained from all participants prior to participation. Patients with FLE were identified through the Alberta Health Services — Calgary Zone Epilepsy Clinic, the Seizure
1288
Table 1 Gender
FLE patient characteristics. Age at scan
Patients with right FLE M 16
Age at seizure onset
Seizure burden
Months since last seizure
Seizure focus
Seizure types
MRI
FDG-PET hypometabolism
Neuro-psychology testing
7
Low
7
Supplementary sensorimotor
GTC, FDC
Not performed
Mild right frontal lobe dysfunction
Unremarkable
Mild right frontal lobe dysfunction
Not performed
Not performed
20
12
Moderate
0
Primary motor
GTC, FDC, FNDC
F
21
18
Moderate
2
GTC
M
24
4
Moderate
8
FNDC
Unremarkable
Not performed
Not performed
M
25
17
Low
32
Anterior frontopolar Primary motor Anterior frontopolar
GTC
Not performed
Not performed
F
29
5
Very high
0
Opercular
FDC
hypometabolism in the inferior frontal region corresponding to MRI abnormality
Mild right frontotemporal dysfunction
M
32
26
Moderate
14
Anterior frontopolar
FNDC
Not performed
Not performed
M
33
28
Low
20
Anterior frontopolar
FDC
Not performed
Not performed
F
36
0
Low
10
GTC, FDC
Not performed
Not performed
M
46
39
Low
13
Not performed
47
28
Low
112
Small vessel ischemic changes Right frontal cavernoma
Not performed
M
Supplementary sensorimotor Anterior frontopolar Supplementary sensorimotor
Small T2 signal hyperintensity in right anterior superior frontal cortex Cortical thickening right anterior insula and inferior lateral frontal love near the operculum CT shows right frontal cystic encephalomalacia Rt posterior fossa arachnoid cyst (incidental) Unremarkable
Not performed
Not performed
Mean ± SD
29.9 ± 10.2 16.7 ± 12.4
19.8
GTC, FDC GTC
K.E. Woodward et al.
F
Focal abnormality in the subcortical superior aspect of the frontal lobe Nonspecific right subcortical frontal T2 hyperintensity Unremarkable
4
Low
6
Supplementary sensorimotor Primary motor Supplementary sensorimotor Dorsolateral
GTC, FDC, FNDC GTC, FDC, FNDC GTC, FNDC
Unremarkable
Not performed
Not performed
Unremarkable
Not performed
Not performed
Unremarkable
Not performed
GTC, FDC
Unremarkable
Small area of gliosis in left primary motor cortex Encephalomalacia (post-traumatic) in left lateral frontal lobe at vertex and in left anteriolateral temporal lobe
Hypometabolism in left perisylvian frontal lobe Not performed
Mild left frontal lobe dysfunction Left frontal lobe dysfunction
M
28
2
Very high
2
M
30
8
High
0
M
39
2
Very high
1
F
55
12
Low
18
Primary motor
GTC, FNDC
F
65
41
Low
30
Primary motor
CP, FNDC
Mean ± SD
39.3 ± 17.5 11.5 ± 15.0
Left anterior temporal hypometabolism
The effect of FLE on motor organization
Patients with left FLE M 19
Not performed
Not performed
9.5
Seizure types are listed as FNDC (focal without dyscognitive features), FDC (focal with dyscognitive features), GTC (generalized tonic-clonic). Seizure burden was evaluated at the time of study as low (seizures every six months or longer), moderate (every one to six months), high (every two to four weeks), or very high (every two weeks or less).
1289
1290 Monitoring Unit of the Foothills Medical Center, and the Comprehensive Epilepsy Program, in Calgary, AB. Recruitment and scanning took place concurrently between May of 2010 and February of 2013. Twenty-three patients were recruited and scanned. Six patients were excluded due to (1) inability to complete the study (one patient), (2) being left-hand dominant, as determined by identifying the dominant writing hand (three patients), and (3) having a change in epilepsy diagnosis from FLE to temporal lobe epilepsy based on subsequent video-EEG monitoring investigations (two patients). This left 17 patients: 11 right FLE (7M, 4F; age = 29.9 ± 10.2 years) and six left FLE (4M, 2F; age = 39.3 ± 17.5 years) (Table 1). The number of patients studied was a sample of convenience based on the available patients. A diagnosis of FLE was confirmed by history, examination, routine EEG, anatomical MR imaging, and in 8 patients, scalp video-EEG monitoring. Ten healthy controls (5M, 5F; age = 33.9 ± 12.7 years) were recruited through word-of-mouth, and had no known neurological or psychiatric disorders. Other exclusion criteria included contraindication to MR imaging [including pregnancy, severe claustrophobia, metallic foreign bodies in the eye, and certain implanted medical devices (e.g., cardiac pacemaker, aneurysm clip)] and previous brain surgery. Two patients underwent surgical resection following the study and pathology showed microdysgenesis in one (left FLE, male, age 30 years) and mild gliosis in the other (right FLE, female, age 29 years). All participants were righthanded.
Behavioral motor testing Participants performed motor tasks by following a visual cue while in the MR scanner. A video projector situated behind the scanner projected visual stimuli onto a screen, which participants viewed using a mirror mounted on the head coil. Participants performed two motor tasks similar to Luria’s Motor Sequences (Luria and Haigh, 1973). The first, a fingertapping task, was used due to its simplicity and ease of use for both healthy participants and those with motor impairments. Participants tapped their fingers in time with a visual cue (contrasting colors alternating at 1-s intervals), and each run consisted of alternating blocks of rest (24 s) and task (12 s), for a total of 3 min and 48 s. The task was performed in separate runs: unimanually with the left and right hand, and bimanually with synchronous tapping. While finger-tapping elicits a brain response predominantly within primary motor regions, more complex tasks elicit strong activity within secondary and tertiary motor regions (Leonard et al., 1988). This recruitment pattern is more representative of daily functions. Therefore, a motor coordination task was also used, which required participants to perform a series of hand movements in time with a visual stimulus (circle = hand held in fist, vertical line = hand held vertically, horizontal line = hand held horizontally) paced at 2-s intervals. Each run consisted of alternating blocks of rest (24 s) and task (24 s) for a total of 5 min. Tasks were performed in separate runs: unimanually with the left and right hand, and bimanually. During the bimanual coordination task, participants performed different hand movements with each hand simultaneously.
K.E. Woodward et al. The order of the six runs was randomized for each participant. During rest periods for both tasks, participants focused on a central fixation cross. Participants practiced the tasks outside of the scanner until they were comfortable performing them. Video recordings of the coordination task were made for subsequent analysis of performance. Participants’ performance was rated blindly using a four point scale (Helmstaedter et al., 1996), ranging from one (no impairment) to four (severe impairment; Supplementary Table 1). A Mann—Whitney-U test was used to identify significant performance differences between patients (right and left FLE together) and controls. Additionally, a Kruskal—Wallis test was used to identify significant performance differences between right FLE, left FLE, and controls. Individual Mann—Whitney-U tests were then conducted between all group pairs, and a Bonferroni correction was applied to correct for multiple comparisons.
fMRI data collection Participants were asked to keep as still as possible, and their head was immobilized using foam cushioning. They could terminate the study at any time during the scan using a squeeze ball placed by their side. MR data were acquired using a 3.0 T GE Discovery MR750 whole body scanner (GE Healthcare, Waukesha, WI) with a receive-only 8-channel phased-array head coil (8 Channel High Resolution Brain Array, distributed by GE Medical Systems). MR images providing BOLD contrast were collected using a gradient-recalled echo, echo planar imaging sequence (voxel dimensions 3.75 × 3.74 × 4 mm, 28 slices, 4 mm slice thickness, 64 × 64 matrix, TE = 30 ms, TR = 1.5 s, flip angle = 65 degrees). In total, six randomized functional MRI scans were completed: one for each finger-tapping task (left, right, bimanual), and one for each coordination task (left, right, bimanual). Participants also underwent T1-weighted multi-slice spoiled gradient echo (28 × 4 mm slices, 128 × 128 matrix, minimum TE, TR = 150 ms, flip angle = 18 degrees) and 3D magnetization-prepared gradient-echo sequences (2 mm slices, 384 × 256 × 112 matrix, preparation time = 500 ms, minimum TE, TR = 8.9 ms, flip angle = 20 degrees) for anatomical registration of the fMRI data.
Preprocessing Preprocessing of image data was performed using FSL [fMRIB Software Library; http://www.fmrib.ox.ac.uk/fsl/] (Jenkinson et al., 2012). These steps included brain extraction using the Brain Extraction Tool (Smith, 2002), correction for interleaved slice timing using Fourier-space time-series phase-shifting, motion correction using MCFLIRT [Motion Correction: FMRIB’s Linear Image Registration Tool] (Jenkinson et al., 2002), spatial smoothing with a 6 mm FWHM Gaussian kernel, high pass temporal filtering (Gaussian-weighted least-squares straight line fitting with sigma = 100.0 s), grand-mean intensity normalization of the entire 4D dataset using a single multiplicative factor, and registration to the common brain template of the Montreal Neurological Institute (MNI152 standard brain, © 1993—2009
The effect of FLE on motor organization Louis Collins, McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University).
First-level analysis Models were created for each task by convolving the binary boxcar function (indicating the timing of task blocks) with the standard hemodynamic response. Models were then used in a time-series General Linear Model analysis using FEAT (Jenkinson et al., 2012) to determine brain regions that were significantly more active during task performance than rest in each participant.
Higher-level analysis Group analyses of parameter estimates were conducted to obtain mean BOLD activity maps for each task in each group. This analysis was carried out in FSL using FEAT, which utilizes the General Linear Mixed Model for higher-level analyses. This process registered each first-level analysis’ parameter estimates to the MNI152 standard brain in order to compute the average response magnitude for each voxel. Specifically, FSL’s FLAME (FMRIB’s Local Analysis of Mixed Effects) was used, which takes the session and participant variability into account, therefore allowing inferences to be made about the wider population and not only about individuals whom participated in the study. Contrasts of parameter estimates were generated to compute mean BOLD activity for each group and between groups and tasks. Specifically, the General Linear Mixed Model was applied by FSL to determine voxels exhibiting a significant difference in response magnitude between groups and tasks. Participant’s coordination performance scores were demeaned and entered as a covariate when computing both the mean and contrast images involving the coordination data. Maps were generated using a corrected cluster significance of p = 0.05, using AlphaSim, which uses Monte Carlo simulations of null distribution image data to estimate family-wise error rates (Ward, 2000).
Laterality indices A laterality index (LI) was used as an indicator of hemispheric dominance (in terms of BOLD activity levels) during task performance. LI was calculated for each individual’s left-handed and right-handed coordination and fingertapping maps, as follows: LI =
1291 they could occur in any cortical region and not just regions commonly thought of as ‘cortical motor areas’. Because the whole-hemisphere was used in the calculations, solely comparing the number of voxels would not quantify changes in BOLD activity levels within each hemisphere, and would just be a measure of brain size. Thus, average Z score was used in the calculation to determine hemispheric dominance during task performance. Control participants LIs were calculated in the same manner as the FLE group that it was being compared to. For example, when controls and right FLE patients were compared using the right-handed coordination task, ‘epileptic’ was the right hemisphere in both groups, and ‘healthy’ was the left hemisphere in both groups. A positive LI value indicated epileptic hemispheric dominance during task performance, whereas a negative LI value indicated healthy hemispheric dominance during task performance. Spearman’s rank correlation analyses were performed between LI and the following factors for each patient group: age at epilepsy diagnosis, years since diagnosis, lifetime seizures (total or generalized tonic-clonic seizures alone), seizures in past year (total or generalized tonic-clonic seizures alone), number of months since the last seizure, and coordination performance score (i.e., one, two, three, or four). Seizure burden factors were determined by retrospectively reviewing the patients’ charts.
Results Behavioral motor testing All participants received a score of one on the unimanual coordination task. On the bimanual coordination task, FLE patients performed significantly worse than control subjects (p = 0.011), in particular the left FLE patients (p = 0.024) (Table 2). The scores for each group were as follows: right FLE: 1.4 ± 0.5; left FLE 2.0 ± 0.9; control 1.0 ± 0.0. There was no significant relationship between motor performance score and length of seizure freedom.
Brain activation During all tasks (right-handed, left-handed, bimanual in both coordination and finger-tapping), control participants showed BOLD activity in cortical regions including
(# voxels epileptic)(avg Z epileptic) − (# voxels healthy)(avg Z healthy) (# voxels epileptic)(avg Z epileptic) + (# voxels healthy)(avg Z healthy)
where ‘epileptic’ was the hemisphere ipsilateral to the seizure focus, and ‘healthy’ was the hemisphere contralateral to the seizure focus. ‘# voxels’ included all voxels in the gray matter of the respective hemisphere, and ‘avg Z’ was the average Z score of all of the voxels in the gray matter of the respective hemisphere. Whole-hemisphere calculations were used because it was hypothesized that if changes in cortical organization were to occur due to epileptic activity,
the primary motor, supplementary motor, prefrontal, posterior parietal, and lateral occipital cortex (Supplementary Figure 1). Regions of BOLD activity in subcortical structures included the putamen, globus pallidus, thalamus, and caudate. Areas of reduced BOLD activity during task performance included the medial prefrontal cortex, posterior
1292
K.E. Woodward et al.
Table 2 Comparison of bimanual coordination performance scores between each of the three groups, analyzed using Kruskal—Wallis (KW) and Mann—Whitney-U (MW) tests. Significant differences (p < 0.05) are shown in bold and marked by an asterisk (*). Comparison
P-value
Bonferroni corrected p-value (when necessary)
FLE vs Controls (MW) All three groups (KW) Right FLE vs Controls (MW) Left FLE vs Controls (MW) Right FLE vs Left FLE (MW)
0.01 (1-tailed) 0.01 0.06 (1-tailed) 0.008 (1-tailed) 0.08 (2-tailed)
0.01* 0.01* 0.17 0.02* 0.45
cingulate, angular gyrus, and posterior insula, all regions that are part of the default mode network. Fig. 1 shows brain regions of right FLE patients where BOLD activity was significantly different than controls during task performance (summarized in Supplementary Table 2). During the left-handed finger-tapping task, right FLE patients showed significantly less BOLD activity in the right middle frontal gyrus, right inferior parietal cortex, and right occipital pole. During the right-handed finger-tapping task, right FLE patients had less BOLD activity in the right and left lateral occipital cortex, right lingual gyrus, and right inferior parietal cortex. During the bimanual finger-tapping task, right FLE patients had less BOLD activity in the right and left frontal pole, right insula, right lateral occipital cortex, right inferior parietal cortex, right superior parietal cortex, right putamen, right thalamus, and precuneus. During the left-handed coordination task, right FLE patients exhibited greater BOLD activity in the left inferior parietal cortex and left precentral gyrus (Fig. 1, Supplementary Table 2). During the right-handed coordination task right FLE patients also had decreased BOLD in the right lateral occipital cortex, precuneus, right superior parietal cortex, and right precentral gyrus. During the bimanual coordination task, these patients showed greater BOLD activity in the left superior and middle temporal gyri, left insular cortex, anterior cingulate, left superior parietal cortex, left superior frontal gyrus, as well as less BOLD activity in the right frontal operculum and right precentral gyrus. Fig. 2 shows brain regions of left FLE patients where BOLD activity was significantly different than controls during task performance (summarized in Supplementary Table 3). During the left-handed finger-tapping task, patients had decreased BOLD activity in the right middle frontal gyrus, right frontal pole, right lateral occipital cortex, and left frontal pole. During the right-handed finger-tapping task, patients had decreased BOLD activity in the left superior parietal lobule. During the bimanual finger-tapping task, patients had decreased BOLD activity in the right lateral occipital cortex, right superior frontal gyrus, right orbitofrontal cortex, right putamen, left frontal pole, left superior parietal lobule, and left precuneus. During the left-handed coordination task, patients had decreased BOLD activity in the right inferior parietal cortex, right middle frontal gyrus, and right putamen (Fig. 2, Supplementary Table 3). During the right-handed coordination task, patients had decreased BOLD activity in the right caudate and left insula. During the bimanual coordination
task, patients had decreased BOLD activity in the right inferior parietal cortex, and increased BOLD activity in the left superior temporal gyrus and left insula.
Relationship between LI and seizure demographics During the right-handed coordination task, control participants had an average LI of −0.02 ± 0.59 (range = −0.90 to 0.94), and an average LI of 0.04 ± 0.40 (range = −0.71 to 0.60) during the left-handed coordination task. During the right-handed finger-tapping task, control participants had an average LI of −0.22 ± 0.39 (range = −0.99 to 0.54), and an average LI of −0.10 ± 0.20 (range = −0.44 to 0.11) during the left-handed finger-tapping task. None of the tasks differed significantly from 0. Patients with right FLE had an average LI of 0.97 ± 1.59 (range = −1.68 to 3.98) during the left-handed coordination task, with five patients falling either above or below the control (‘normal’) LI range, and an average LI of 0.22 ± 0.55 (range = −0.55 to 0.99) during the left-handed finger-tapping task. Patients with left FLE had an average LI of 0.10 ± 0.30 (range = −0.40 to 0.41) during the righthanded coordination task, and an average LI of 0.55 ± 0.43 (range = 0.20 to 0.92) during the right-handed finger-tapping task. There was no significant difference between LIs of right or left FLE patients and controls. All participants received a performance score of one on the unimanual coordination tasks. Therefore, when determining the effects of motor impairments on LI, only the bimanual coordination performances scores were compared. Patients with a score of one (n = 9), two (n = 6), and three (n = 3) had an average LI of 0.99 ± 1.33, 0.48 ± 1.52, and −0.23 ± 0.24, respectively. No patients received a score of 4. There was no significant difference between groups. In FLE patients, a positive correlation was seen between the number of months since the last seizure, and the LI of the coordination task that employed the epileptic hemisphere (i.e., left-handed coordination task for right FLE patients, rS = 0.779; right-handed coordination task for left FLE patients, rS = 0.943; Fig. 3). No correlation was observed between LI and the coordination task that would typically employ the healthy hemisphere, or between LI and the finger-tapping tasks. Fig. 3 also shows the relationship between the duration of seizure freedom and brain activation during the unimanual coordination task. Specifically, with longer periods of seizure freedom, right FLE patients showed less BOLD signal in the left sensorimotor cortex, and
The effect of FLE on motor organization
1293
Figure 1 Comparison of brain activation between right FLE patients and controls during finger-tapping and coordination tasks. Areas of red/yellow indicate regions that were more active in right FLE patients compared to controls during task performance. Areas of blue/light blue indicate regions that were less active in right FLE patients compared to controls during task performance.
left and right inferior parietal cortex, and left FLE patients showed less BOLD signal in the right sensorimotor cortex, and left and right occipital cortex (summarized in Supplementary Table 4).
Discussion This was the first fMRI study to compare motor activation between a group of adult FLE patients and healthy controls.
Right FLE patients had significant decreases in activation of the epileptic hemisphere, and significant increases in activation of the healthy hemisphere during task performance. Results were less conclusive in the left FLE patients, possibly due to a limited number of patients. Additionally, in both patient groups there was a positive correlation between sensorimotor cortex activation in the epileptic hemisphere and the duration of seizure freedom prior to the study.
1294
K.E. Woodward et al.
Figure 2 Comparison of brain activation between left FLE patients and controls during finger-tapping and coordination tasks. Areas of red/yellow indicate regions that were more active in left FLE patients compared to controls during task performance. Areas of blue/light blue indicate regions that were less active in left FLE patients compared to controls during task performance.
Finger-tapping tasks During unimanual finger-tapping tasks, both patients groups showed decreased BOLD activity compared to controls (Figs. 1 and 2). One region common to both group comparisons was the posterior parietal cortex of the epileptic hemisphere, solely when performing tasks that primarily taxed the epileptic hemisphere (i.e., left-handed task for
right FLE patients, right-handed task for left FLE patients). This may have been due to detrimental effects of seizure propagation from the ipsilateral seizure focus, causing patients to rely less on the epileptic hemisphere. Indeed, seizure propagation paths form strong neuronal networks due to synchronous and repetitive activity, which could disrupt normal networks in the brain, such as the motor network. The posterior parietal cortex is part of the motor
The effect of FLE on motor organization
1295
Figure 3 Left: relationship between LIs of the left-handed coordination task and months since last seizure in right FLE patients (rs = 0.779), and LIs of the right-handed coordination task and months since last seizure in left FLE patients (rs = 0.943). Increasingly positive LI values indicate greater dominance of the epileptic hemisphere during task performance. Right: relationship between duration of seizure freedom and brain activation during the unimanual coordination task. Images indicate brain regions that were more active (red) or less active (blue) as a function of the duration of seizure freedom for the left-handed coordination task in right FLE patients (upper row) and the right-handed coordination task in left FLE patients (lower row).
network, and is pivotal for synchronizing hand movements with external stimuli, as well as integrating sensory and motor information (Iacoboni, 2006). While we did not observe motor impairments during unimanual finger-tapping tasks, decreased BOLD activity in the posterior parietal cortex may in part be responsible for daily motor impairments seen in FLE patients. Similar to unimanual finger-tapping tasks, the bimanual finger-tapping task was associated with decreased BOLD activity in the posterior parietal cortex of the epileptic hemisphere in both patient groups (Figs. 1 and 2). Additionally, during the bimanual tapping task both patient groups had less BOLD activity in brain regions that are part of the default mode network. The default mode network is primarily activated in the absence of external stimuli, and brain activity is increasingly suppressed during more complex tasks (McKiernan et al., 2003). This suggests that patients focused more on external stimuli than did controls, possibly to overcome motor impairments in order to perform the task fluently.
Coordination tasks During unimanual coordination tasks, right FLE patients had increased BOLD signal in motor regions of the healthy hemisphere during the task that taxed the epileptic hemisphere (left-handed coordination, Fig. 1). This suggests that normal recruitment of the epileptic hemisphere was insufficient to perform the left-handed task, and that additional, healthy hemisphere brain regions were recruited as a means of compensation. However, this compensatory activation was not as effective as normal activation patterns because the patient group did not perform as well as the control group (Table 2). This suggests that while healthy hemisphere brain regions
were necessary for patients to complete the task, they were not as effective in eliciting correct hand maneuvers as the cortex that was originally responsible for such movements. Furthermore, right FLE patients had decreased BOLD signal in motor regions of the epileptic hemisphere during the task that primarily taxed the healthy hemisphere (righthanded coordination). This suggests that the right-handed task rendered compensatory activity unnecessary for adequate motor performance, as the epileptic hemisphere was not being recruited as heavily. The comparison of unimanual coordination tasks between left FLE and controls produced more inconsistent results with no clear trends. We suspect that this variability may be due to the limited number of left FLE patients included in the study. During the bimanual coordination task, both patient groups had regions of increased left hemisphere BOLD activity (Figs. 1 and 2). This suggests that these changes in activation might not be related to seizure focus lateralization, but rather hemispheric dominance. All participants were right-handed; therefore, these findings suggest increased reliance on regions in the left, or dominant, hemisphere when compared to controls. Dominant hemisphere connections are generally stronger because of increased use (Chen et al., 1997; Kim et al., 1993); therefore, they may be less susceptible to detrimental seizure activity. If non-dominant hemisphere connections were impacted by seizure activity, this may cause patients to rely more on the dominant hemisphere during motor task performance, as was seen in the bimanual coordination task. During bimanual coordination, increased BOLD signal was seen in the left insula and superior temporal gyrus in both FLE groups (Figs. 1 and 2). Right FLE patients also had increased BOLD activity in the left anterior cingulate. All of these regions have been implicated in error-detection, and show increased BOLD signal during complex motor
1296 tasks when compared to simpler tasks (Carter et al., 1998; Debaere et al., 2004; Ullsperger and von Cramon, 2003; Wendelken et al., 2009). This suggests that patients made more errors, presumably due to the difficulty of the bimanual coordination task. Indeed, patients had significantly poorer performance scores on the bimanual coordination task when compared to controls (Table 2). Additionally, increased BOLD activity was observed in the superior temporal gyrus, which is activated during sequence processing tasks (Ullen et al., 2005), such as the coordination task in this study. However, less activation was seen with increased practice (Jantzen et al., 2002), again suggesting that patients had not mastered the task as well as control participants.
Relationship between LI and period of seizure freedom During coordination tasks that taxed the epileptic hemisphere, patients with recent seizures had increased BOLD activity in the sensorimotor cortex of the healthy hemisphere, whereas patients with longer periods of seizure freedom had less BOLD activity in this region (Fig. 3). This suggests that propagation of abnormal neuronal activity, due to recent seizures, adversely affected the epileptic hemisphere, thereby causing patients to rely more on sensorimotor cortex of the healthy hemisphere. These findings are supported by studies of motor recovery following stroke and epilepsy surgery (multiple subpial transection of the seizure focus). Patients with strokes or seizure foci in primary motor cortex initially had increased BOLD activity in the healthy hemisphere when using the affected hand, which persisted over many weeks; however, there was an eventual return to pre-stroke, or pre-surgical activation patterns (Chollet et al., 1991; Moo et al., 2002; Feydy et al., 2002). Furthermore, patients demonstrated no changes in brain activation over time when testing the unaffected hand (Chollet et al., 1991; Moo et al., 2002). This is similar to the results of our study, in which no correlation was seen between LI of the task using the unaffected hand and period of seizure freedom.
Pathophysiology of changes in motor cortex organization Given the relatively rapid changes in brain activation described in the previous section, the reversibility of these changes, and distance of changes from the lesion site, it is likely that pre-existing but dormant connections were in place to facilitate these changes (Feydy et al., 2002; Moo et al., 2002). Indeed, horizontal cortico-cortical connections are known to exist between primary motor cortex and other motor regions, both within and between hemispheres (Castro-Alamancos and Borrel, 1995; Feydy et al., 2002). If such connections became less active following a lesion or recurrent seizures, this could result in disinhibition, thereby allowing recruitment of additional sites to maintain normal motor function. Indeed, patients with more frequent seizures have impaired cortico-cortical inhibition as measured by transcranial magnetic stimulation (Labyt et al., 2007). However, once normal conditions are reinstated, or
K.E. Woodward et al. perhaps long periods of seizure freedom ensue, inhibition can be re-exerted upon compensatory cortical regions and normal activation patterns return (Feydy et al., 2002; Moo et al., 2002). Thus, recent seizures could initially increase recruitment of healthy hemisphere regions during coordination tasks, which would return back to baseline following longer periods of seizure freedom (Fig. 3). Previous studies using rat models of epilepsy have also suggested that disinhibition may underlie seizure-induced changes in cortical motor map representations (CastroAlamancos and Borrel, 1995; van Rooyen et al., 2006; Young et al., 2009). Both decreases in GABAergic and increases in glutamatergic transmission, which occur in humans with epilepsy, have been shown to influence the recruitment of additional motor regions through pre-existing corticocortical connections (Behr et al., 2000; Hess et al., 1996; van Rooyen et al., 2006). Additionally, in rats, repeated kindling near the sensorimotor cortex leads to impaired interictal behavioral motor performance (Henry et al., 2008). Such altered motor performance was associated with more efficacious synapses, providing further evidence for disinhibition as an underlying mechanism of impaired motor function. This is shown in our results as patients with the worst performance scores had the lowest average LI values, indicating greater reliance (possibly through disinhibition) on the healthy hemisphere. As performance scores improved, LI values increased to indicate greater reliance on the epileptic hemisphere, conceivably by establishing a more normal pattern of BOLD activity through a re-exertion of inhibition on healthy hemisphere connections.
Limitations of study We did not study a large number of participants, due to the limited number of FLE patients who were candidates for the study. Increasing the number of participants might further clarify our results, as we would be able to separate patients into FLE patient subgroups based on the specific location of the seizure focus within the frontal lobes, amongst other patient-specific factors. Indeed, the location of the seizure foci within the frontal lobe (Koudijs et al., 2010) as well as seizure frequency, severity, and duration (Upton and Thompson, 1996) affect the results of task-based fMRI. However, we attempted to account for patient heterogeneity by determining the effect of a variety of patient-specific factors on LIs. When we did this, we found a positive correlation between the number of months since the last seizure and the LI for the coordination task in both patient groups. Despite the average age difference between groups, there was no statistically significant difference in age. A previous study that demonstrated differences in motor activity due to age examined two groups of patients with average ages of 23 and 66 for the younger and older groups, respectively, a difference of 43 years (Riecker et al., 2006). This is much larger then the maximum difference of 10 years in our study. There were also small differences in the number of males and female in each group; however, other studies of gender differences in motor activation during task performance have shown that there is no difference in activation patterns between males and females (Ward and Frackowiak, 2003).
The effect of FLE on motor organization Patients were taking a variety of anticonvulsant medications in combination or isolation. Previous studies have shown that these drugs can decrease BOLD activity during task-based fMRI studies (Jansen et al., 2006; Szaflarski and Allendorfer, 2012), or have no significant effect on BOLD activity (Braakman et al., 2013; Yasuda et al., 2013). We also confirmed that anticonvulsant medications have no significant effect on the BOLD signal during task-based motor fMRI (unpublished observations), which suggests that the results of the present study were not likely related to medication effect.
Conclusions This was the first known study to examine motor fMRI activation in a group of adult patients with FLE. In general, patients showed significantly greater BOLD activity in brain regions contralateral to the seizure focus, and significantly less BOLD activity in brain regions ipsilateral to the seizure focus while performing a motor task. Additionally, patients with recent seizures had an initial reliance on the healthy hemisphere, and an eventual recovery of epileptic hemisphere BOLD activity with longer periods of seizure freedom. These results suggest that changes in cortical activation may be an underlying mechanism of motor deficits observed in FLE patients, and suggest the potential for longer periods of seizure freedom to reduce these motor deficits.
Conflict of interest None of the authors have 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.
Acknowledgements This work was supported by CIHR (MOP-230809). We also thank Daniel J Pittman and Aaron Spring for critically reviewing the manuscript.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.eplepsyres.2014.06.015.
References Behr, J., Heinemann, U., Mody, I., 2000. Glutamate receptor activation in the kindled dentate gyrus. Epilepsia 41 (Suppl. 6), S100—S103. Braakman, H.M., Vaessen, M.J., Jansen, J.F., Debeij-van Hall, M.H., de Louw, A., Hofman, P.A., Vles, J.S., Aldenkamp, A.P., Backes, W.H., 2013. Frontal lobe connectivity and cognitive impairment in pediatric frontal lobe epilepsy. Epilepsia 54, 446—454. Branco, D.M., Coelho, T.M., Branco, B.M., Schmidt, L., Calcagnotto, M.E., Portuguez, M., Neto, E.P., Paglioli, E., Palmini, A., Lima, J.V., Da Costa, J.C., 2003. Functional variability of the
1297 human cortical motor map: electrical stimulation findings in perirolandic epilepsy surgery. J. Clin. Neurophysiol. 20, 17—25. Carter, C.S., Braver, T.S., Barch, D.M., Botvinick, M.M., Noll, D., Cohen, J.D., 1998. Anterior cingulate cortex, error detection, and the online monitoring of performance. Science 280, 747—749. Castro-Alamancos, M.A., Borrel, J., 1995. Functional recovery of forelimb response capacity after forelimb primary motor cortex damage in the rat is due to the reorganization of adjacent areas of cortex. Neuroscience 68, 793—805. Chen, R., Gerloff, C., Hallett, M., Cohen, L.G., 1997. Involvement of the ipsilateral motor cortex in finger movements of different complexities. Ann. Neurol. 41, 247—254. Chollet, F., DiPiero, V., Wise, R.J., Brooks, D.J., Dolan, R.J., Frackowiak, R.S., 1991. The functional anatomy of motor recovery after stroke in humans: a study with positron emission tomography. Ann. Neurol. 29, 63—71. Debaere, F., Wenderoth, N., Sunaert, S., Van Hecke, P., Swinnen, S.P., 2004. Changes in brain activation during the acquisition of a new bimanual coodination task. Neuropsychologia 42, 855—867. Exner, C., Boucsein, K., Lange, C., Winter, H., Weniger, G., Steinhoff, B.J., Irle, E., 2002. Neuropsychological performance in frontal lobe epilepsy. Seizure: J. Br. Epilepsy Assoc. 11, 20—32. Feydy, A., Carlier, R., Roby-Brami, A., Bussel, B., Cazalis, F., Pierot, L., Burnod, Y., Maier, M.A., 2002. Longitudinal study of motor recovery after stroke: recruitment and focusing of brain activation. Stroke: J. Cereb. Circ. 33, 1610—1617. Gaillard, W.D., Berl, M.M., Duke, E.S., Ritzl, E., Miranda, S., Liew, C., Finegersh, A., Martinez, A., Dustin, I., Sato, S., Theodore, W.H., 2011. fMRI language dominance and FDG-PET hypometabolism. Neurology 76, 1322—1329. Hamer, H.M., Reis, J., Mueller, H.H., Knake, S., Overhof, M., Oertel, W.H., Rosenow, F., 2005. Motor cortex excitability in focal epilepsies not including the primary motor area — a TMS study. Brain: J. Neurol. 128, 811—818. Helmstaedter, C., Kemper, B., Elger, C.E., 1996. Neuropsychological aspects of frontal lobe epilepsy. Neuropsychologia 34, 399—406. Henry, L.C., Goertzen, C.D., Lee, A., Teskey, G.C., 2008. Repeated seizures lead to altered skilled behaviour and are associated with more highly efficacious excitatory synapses. Eur. J. Neurosci. 27, 2165—2176. Hess, G., Aizenman, C.D., Donoghue, J.P., 1996. Conditions for the induction of long-term potentiation in layer II/III horizontal connections of the rat motor cortex. J. Neurophysiol. 75, 1765—1778. Iacoboni, M., 2006. Visuo-motor integration and control in the human posterior parietal cortex: evidence from TMS and fMRI. Neuropsychologia 44, 2691—2699. Jansen, J.F., Aldenkamp, A.P., Marian Majoie, H.J., Reijs, R.P., de Krom, M.C., Hofman, P.A., Eline Kooi, M., Nicolay, K., Backes, W.H., 2006. Functional MRI reveals declined prefrontal cortex activation in patients with epilepsy on topiramate therapy. Epilepsy Behav. 9, 181—185. Jantzen, K.J., Steinberg, F.L., Kelso, J.A., 2002. Practicedependent modulation of neural activity during human sensorimotor coordination: a functional magnetic resonance imaging study. Neurosci. Lett. 332, 205—209. Janszky, J., Ebner, A., Kruse, B., Mertends, M., Jokeit, H., Seitz, R.J., Witte, O.W., Tuxhorn, I., Woermann, F.G., 2003. Functional organization of the brain with malformations of cortical development. Ann. Neurol. 53, 759—767. Janszky, J., Mertens, M., Janszky, I., Ebner, A., Woermann, F.G., 2006. Left-sided interictal epileptic activity induces shift of language lateralization in temporal lobe epilepsy: an fMRI study. Epilepsia 47, 921—927. Jenkinson, M., Bannister, P., Brady, M., Smith, S., 2002. Improved optimization for the robust and accurate linear registration and motion correction of brain images. NeuroImage 17, 825—841.
1298 Jenkinson, M., Beckmann, C.F., Behrens, T.E., Woolrich, M.W., Smith, S.M., 2012. FSL. NeuroImage 62, 782—790. Kim, S.G., Ashe, J., Hendrich, K., Ellermann, J.M., Merkle, H., Ugurbil, K., Georgopoulos, A.P., 1993. Functional magnetic resonance imaging of motor cortex: hemispheric asymmetry and handedness. Science 261, 615—617. Koudijs, S.M., Leijten, F.S., Ramsey, N.F., van Nieuwenhuizen, O., Braun, K.P., 2010. Lateralization of motor innervation in children with intractable focal epilepsy — a TMS and fMRI study. Epilepsy Res. 90, 140—150. Labyt, E., Houdayer, E., Cassim, F., Bourriez, J.L., Derambure, P., Devanne, H., 2007. Motor representation areas in epileptic patients with focal motor seizures: a TMS study. Epilepsy Res. 75, 197—205. Leonard, G., Milner, B., Jones, L., 1988. Performance on unimanual and bimanual tapping tasks by patients with lesions of the frontal or temporal lobe. Neuropsychologia 26, 79—91. Luria, A.R., Haigh, B., 1973. The working brain. (London) Allen Lane the Penguin Press, Penguin. Macdonell, R.A., Jackson, G.D., Curatolo, J.M., Abbott, D.F., Berkovic, S.F., Carey, L.M., Syngeniotin, A., Fabinyi, G.C., Scheffer, I.E., 1999. Motor cortex localization using functional MRI and transcranial magnetic stimulation. Neurology 53, 1462—1467. McKiernan, K.A., Kaufman, J.N., Kucera-Thompson, J., Binder, J.R., 2003. A parametric manipulation of factors affecting task-induced deactivation in functional neuroimaging. J. Cogn. Neurosci. 15, 394—408. Moo, L.R., Slotnick, S.D., Krauss, G., Hart, J., 2002. A prospective study of motor recovery following multiple subpial transections. Neuroreport 13, 665—669. Muller, R.A., Rothermel, R.D., Behen, M.E., Muzik, O., Chakraborty, P.K., Chugani, H.T., 1997. Plasticity of motor organization in children and adults. Neuroreport 8, 3103—3108. Riecker, A., Groschel, K., Ackermann, H., Steinbrink, C., Witte, O., Kastrup, A., 2006. Functional significance of age-related differences in motor activation patterns. NeuroImage 32, 1345—1354. Smith, S.M., 2002. Fast robust automated brain extraction. Hum. Brain Mapp. 17, 143—155. Stoeckel, M.C., Kleinschmidt, A., Ebner, A., Witte, O.W., Seitz, R.J., 2002. Reorganization of motor representation in a patient with epilepsia partialis continua as shown by [O15]-labeled butanol positron emission tomography and functional magnetic resonance imaging. J. Neuroimaging: Off. J. Am. Soc. Neuroimag. 12, 276—281.
K.E. Woodward et al. Szaflarski, J.P., Allendorfer, J.B., 2012. Topiramate and its effect on fMRI of language in patients with right or left temporal lobe epilepsy. Epilepsy Behav. 24, 74—80. Tellez-Zenteno, J.F., Dhar, R., Wiebe, S., 2005. Long-term seizure outcomes following epilepsy surgery: a systematic review and meta-analysis. Brain: J. Neurol. 128, 1188—1198. Ullen, F., Bengtsson, S.L., Ehrsson, H.H., Forssberg, H., 2005. Neural control of rhythmic sequences. Ann. N Y Acad. Sci. 1060, 368—376. Ullsperger, M., von Cramon, D.Y., 2003. Error monitoring using external feedback: specific roles of the habenular complex, the reward system, and the cingulate motor area revealed by functional magnetic resonance imaging. J. Neurosci.: Off. J. Soc. Neurosci. 23, 4308—4314. Upton, D., Thompson, P.J., 1997. Age at onset and neuropsychological function in frontal lobe epilepsy. Epilepsia 38, 1103—1113. Upton, D., Thompson, P.J., 1996. General neuropsychological characteristics of frontal lobe epilepsy. Epilepsy Res. 23, 169—177. van Rooyen, F., Young, N.A., Larson, S.E., Teskey, G.C., 2006. Hippocampal kindling leads to motor map expansion. Epilepsia 47, 1383—1391. Ward, B.D., 2000. AlphaSim: estimate statistical significance via Monte Carlo simulation. Biophysics Research Institute, Medical College of Wisconsin. Ward, N.S., Frackowiak, R.S.J., 2003. Age-related changes in the neural correlates of motor performance. Brain 126, 873—888. Wendelken, C., Ditterich, J., Bunge, S.A., Carter, C.S., 2009. Stimulus and response conflict processing during perceptual decision making. Cogn. Affect. Behav. Neurosci. 9, 434—447. Woodward, K.E., Gaxiola-Valdez, I., Goodyear, B.G., Federico, P., 2014. Frontal lobe epilepsy alters funtional connections within the brain’s motor network: a resting-state fMRI study. Brain Connect. 4, 91—99. Yasuda, C.L., Centeno, M., Vollmar, C., Stretton, J., Symms, M., Cendes, F., Mehta, M.A., Thompson, P., Duncan, J.S., Koepp, M.J., 2013. The effect of topiramate on cognitive fMRI. Epilepsy Res.. Young, N.A., Vuong, J., Ozen, L.J., Flynn, C., Teskey, G.C., 2009. Motor map expansion in the pilocarpine model of temporal lobe epilepsy is dependent on seizure severity and rat strain. Exp. Neurol. 217, 421—428.
