Cognitive Phenotypes in Juvenile Myoclonic Epilepsy
Abstract number :
2.321
Submission category :
11. Behavior/Neuropsychology/Language / 11B. Pediatrics
Year :
2022
Submission ID :
2204640
Source :
www.aesnet.org
Presentation date :
12/4/2022 12:00:00 PM
Published date :
Nov 22, 2022, 05:26 AM
Authors :
Aaron Struck, MD – University of Wisconsin School of Medicine and Public Health; Camille Garcia-Ramos, PhD – Neurology – University of Wisconsin School of Medicine and Public Health; Vivek Prabhakaran, MD, PhD – Radiology – University of Wisconsin School of Medicine and Public Health; Dace Almane, MS – Neurology – University of Wisconsin School of Medicine and Public Health; Veena Nair, PhD – Radiology – University of Wisconsin School of Medicine and Public Health; Jana Jones, PhD – Neurology – University of Wisconsin School of Medicine and Public Health; Bruce Hermann, PhD – Neurology – University of Wisconsin School of Medicine and Public Health
Rationale: Patients with epilepsy demonstrate considerable variability in their cognitive status, even within discrete epilepsy syndromes. Application of unsupervised machine learning procedures has identified distinct latent cognitive subgroups or phenotypes characterized by diverse patterns of performance ranging from intact to generalized to focally impaired. These patterns have been demonstrated within pediatric and adult temporal lobe epilepsy, frontal lobe epilepsy, and children with focal and generalized new onset epilepsies. Here we examine the presence and nature of cognitive phenotypes and their correlates among the initial participants in the Juvenile Myoclonic Epilepsy Connectome Project (JMECP).
Methods: Twenty-three patients with JME (mean age = 17.4, sd=2.8, 14 females) and 33 controls (mean age= 19.4, sd=4.1, 18 females) were administered a comprehensive neuropsychological battery assessing intelligence (verbal, nonverbal), language (object naming, fluency), immediate and delayed verbal memory (list learning), executive function (set shifting, response inhibition, problem solving), attention (inattention, impulsivity) and psychomotor speed (dominant, nondominant). All scores were age adjusted. The control and JME groups were combined and the cognitive measures were analyzed by unsupervised machine learning analytics to identify latent cognitive groups. Identified clusters were examined in regard to the proportion of JME and control participants within each cluster and then explored was the association of JME cluster membership with sociodemographic, clinical epilepsy, and morphological and functional imaging network features.
Results: Both K-means and hierarchal clustering approaches identified two cognitive clusters (intact, abnormal). There was a significant difference in the proportion of JME and control participants in the two clusters (X2=8.6, df=1, p=0.003) with 33.3% and 66.7% of the JME and control participants respectively in the intact cluster, compared to 77.3% and 26.5% of the JME and control participants respectively in the impaired cluster. The intact versus impaired JME clusters did not differ significantly in age of onset of diagnosis, duration of epilepsy, number of antiseizure medications, handedness, or time since last seizure (all p’s > 0.20). There was a significant difference in IQ metrics between the abnormal (VIQ=88.1 and PIQ= 90.9) and intact JME clusters (VIQ=111.2 and PIQ=112.7) (p’s < 0.05). A cluster characterized by prominent dysexecutive function in the JME cohort was not identified.
Behavior