Authors :
Presenting Author: Gadi Miron, MD – Charité University Hospital
Paul Manuel Müller, MSC – Computational Neurology, Department of Neurology – Charité; Martin Holtkamp, Professor – Epilepsy-Center Berlin-Brandenburg, Department of Neurology – Charité University Hospital; Christian Meisel, Dr – Computational Neurology, Department of Neurology, – Charité University Hospital; Claudia Chien, PhD – Charité University Hospital
Rationale: A new taxonomy for classification of cognitive performance in patients with temporal lobe epilepsy (TLE) has been proposed, including minimal, focal (single domain), and generalised (multi domain) cognitive impairment phenotypes.
1 Prior neuroimaging studies in TLE have established widespread patterns of brain atrophy, however the relationship between cognitive phenotypes and atrophy patterns remains unclear. The aims of this study are (1) to evaluate cortical atrophy patterns in a large TLE patient cohort, (2) to relate them to cognitive phenotypes, and (3) to determine their potential utility for post-surgical cognitive outcome prediction.
Methods: Patients with drug resistant TLE underwent neuropsychological testing and were characterized into three distinct cognitive phenotypes according to IC-CODE criteria.
1 ANOVA was performed to compare cortical thickness measures between patients and age- and sex-matched healthy controls from the Human Connectome Project (HCs). To ensure robustness of results, Bonferroni multiple correction was performed. In patients that underwent epilepsy surgery, logistic regression models utilizing cortical thickness measures were constructed to predict post-surgical cognitive improvement. The performance of these models was assessed using a cross-validation approach.
Results: The study included 124 patients (63 females, mean age 36.0±12.0 years) and 117 age and sex matched healthy controls (63 females, 36.1±12.0 years). Patients with generalized cognitive impairment (n=66, 53.2%) exhibited the most widespread structural changes with 28 cortical regions significantly different from healthy controls, followed by focal impaired patients (n=37, 29.8%) with 13 regions, and minimally impaired patients (n=21, 16.9%) with two affected cortical regions (Figure 1). For generalised and focal impaired patient groups, eight cortical regions overlapped. Regions affected included bilateral anterior cingulate and medial prefrontal cortex, medial and lateral temporal regions, and bilateral auditory association cortices. In sixty-nine (35 females, age 33.6 ± 18.0) patients that underwent surgery, a logistic regression model based on affected ROIs predicted post-surgical verbal memory improvement with an area under the receiving operating curve of 0.69 ±0.15.
Conclusions: Our findings demonstrate a differential pattern of brain atrophy across different cognitive phenotypes in TLE. Specifically, we report distinct cortical regions to be significantly altered compared to age- and sex-matched HCs in each cognitive profile. We show for the first time that cortical thickness patterns associated with cognitive profiles has potential for prediction of post-surgical verbal memory.
Reference:
1 McDonald CR, Busch RM, Reyes A, et al. Development and application of the International Classification of Cognitive Disorders in Epilepsy (IC-CoDE): Initial results from a multi-center study of adults with temporal lobe epilepsy.
Neuropsychology. 2023;37(3):301-314
Funding: Gadi Miron is supported by the Research Training Fellowship grant from the European Academy of Neurology.