Motor hyperactivation during cognitive tasks: An endophenotype of juvenile myoclonic epilepsy

Abstract

Objective: Juvenile myoclonic epilepsy (JME) is the most common genetic generalized epilepsy syndrome. Myoclonus may relate to motor system hyperexcitability and can be provoked by cognitive activities. To aid genetic mapping in complex neuropsychiatric disorders, recent research has utilized imaging intermediate phenotypes (endophenotypes). Here, we aimed to (a) characterize activation profiles of the motor system during different cognitive tasks in patients with JME and their unaffected siblings, and (b) validate those as endophenotypes of JME.

Methods: This prospective cross-sectional investigation included 32 patients with JME, 12 unaffected siblings, and 26 controls, comparable for age, sex, handedness, language laterality, neuropsychological performance, and anxiety and depression scores. We investigated patterns of motor system activation during episodic memory encoding and verb generation functional magnetic resonance imaging (fMRI) tasks.

Results: During both tasks, patients and unaffected siblings showed increased activation of motor system areas compared to controls. Effects were more prominent during memory encoding, which entailed hand motion via joystick responses. Subgroup analyses identified stronger activation of the motor cortex in JME patients with ongoing seizures compared to seizure-free patients. Receiver-operating characteristic curves, based on measures of motor activation, accurately discriminated both patients with JME and their siblings from healthy controls (area under the curve: 0.75 and 0.77, for JME and a combined patient-sibling group against controls, respectively; P < .005).

Significance: Motor system hyperactivation represents a cognitive, domain-independent endophenotype of JME. We propose measures of motor system activation as quantitative traits for future genetic imaging studies in this syndrome.

Keywords: cognition; endophenotype; fMRI; juvenile myoclonic epilepsy; motor system.

FIGURE 1

Activation maps for memory and language functional magnetic resonance imaging (fMRI). The figure shows whole‐brain activation maps, obtained via one‐sample t tests across all subjects, for the effect of encoding pictures, words, and faces (left‐hand side) during the memory fMRI task, and for word repetition and generation during the expressive language fMRI task (right‐hand side). Memory‐associated activation in frontal lobe areas is left lateralized for word encoding, bilateral for picture encoding, and right lateralized for face encoding. Fronto‐temporo‐parietal language fMRI activation is left lateralized, and effects in the left middle and inferior temporal gyrus are more marked during the verb generation condition. Conjunction analyses represent multi‐dimensional equivalent of one‐sided t tests, ³⁰ and highlight consistent effects across task conditions. All activation maps are thresholded at P < .05, family‐wise error (FWE) corrected for multiple comparisons across the whole brain

FIGURE 2

Group comparisons for memory fMRI activation. Across item categories, the figure shows areas of enhanced activation for comparisons of patients with juvenile myoclonic epilepsy (JME) against controls (JME > CTR, panel A), patients with ongoing seizures against those seizure‐free (NSz free > Sz free, panel B), and JME siblings against controls (SIB > CTR, panel E). Comparison of (a) JME patients with ongoing seizures against controls and (b) seizure‐free JME patients against controls is also provided for completeness (panels C and D, respectively). Conjunction analysis ³⁰ identified shared areas of hyperactivation in patients and siblings (panel F). Comparisons for motor system and remainder whole‐brain effects are shown with different color scales (orange‐yellow scale for motor regions, red scale for the remainder brain areas). P‐values for activation differences within the motor system were corrected for family‐wise error rate using 12‐mm diameter spherical regions of interest centered on local maxima. “LH/RH” refer to sagittal sections of the left/right hemisphere. Color bars reflect z‐score scales. MNI coordinates and statistical details are provided in Table S2

FIGURE 3

Group comparisons for language functional magnetic resonance imaging (fMRI) activation. Across language task conditions, the figure displays areas of enhanced activation for comparisons of patients with juvenile myoclonic epilepsy (JME) against controls (JME > CTR, panel A), patients with ongoing seizures against those seizure‐free (NSz free > Sz free, panel B), and JME siblings against controls (SIB > CTR, panel E). Comparison of (a) JME patients with ongoing seizures against controls and (b) seizure‐free JME patients against controls are also provided for completeness (panels C and D, respectively). Conjunction analysis ³⁰ identified shared areas of hyperactivation in patients and siblings (panel F). Comparisons for motor system and remainder whole‐brain effects are shown with different color scales (orange‐yellow scale for motor regions, red scale for the remainder brain areas). P‐values for motor system activation differences were corrected for family‐wise error rate using 12‐mm diameter spherical regions of interest centered on local maxima. “LH” refers to a sagittal section of the left hemisphere. Color bars reflect z‐score scales. Montreal Neurological Institute (MNI) coordinates and statistical details are provided in Table S3

FIGURE 4

Group comparisons for combined activation model across functional magnetic resonance imaging (fMRI) tasks. Pooling across all memory and language conditions, the figure displays areas of enhanced activation for comparisons of patients with juvenile myoclonic epilepsy (JME) against controls (JME > CTR, panel A), patients with ongoing seizures against those seizure‐free (NSz free > Sz free, panel B), and JME siblings against controls (SIB > CTR, panel E). Comparison of (a) JME patients with ongoing seizures against controls and (b) seizure‐free JME patients against controls are also provided for completeness (panels C and D, respectively). Conjunction analysis ³⁰ identified shared areas of hyperactivation in patients and siblings (panel F). Comparisons for motor and remainder whole‐brain effects are shown with different color scales (orange‐yellow scale for motor regions, red scale for the remainder brain areas). P‐values for motor system activation differences were corrected for family‐wise error rate using 12‐mm diameter spherical regions of interest centered on local maxima. “LH” refers to a sagittal section of the left hemisphere. Color bars reflect z‐score scales. MNI coordinates and statistical details are provided in Table S4

FIGURE 5

Receiver‐operating characteristic (ROC) curve analyses on measures of motor activation. The figure shows ROC curves probing usefulness of motor system activation metrics in discriminating patients with juvenile myoclonic epilepsy (JME) from controls (left panel), and JME patients and their siblings, considered as a unitary group, from controls (right panel). Separate analyses were conducted for motor activation during memory and language functional magnetic resonance imaging (fMRI) tasks (orange and beige curves, respectively), whereas a third analysis employed parameter estimates derived from a combined fMRI model, averaging across all cognitive conditions (dark red curve). Individual discrimination of both JME patients and siblings from controls was significant in all analyses (all P < .02), with measures derived from the composite model leading to slightly higher classification accuracy (area under the curve [AUC] 0.75/0.77, for individual discrimination of JME/combined JME‐sibling group from controls, respectively; both P < .005). Full statistical details are provided in section 3.6