(K-409) Comparing Resting-State fMRI Connectivity and Wada Test Memory Lateralization in Epileptic Patients

Introduction: Drug-resistant epilepsy is a harmful condition typically treated with surgical resection of the seizure onset zone [1]. This process, however, requires accurate, preoperational localization of eloquent regions such as those responsible for language and memory to help the surgeon understand which areas should not be resected [2]. The Wada test is a standard procedure used to lateralize language and memory function but is an invasive procedure with risks of complications such as encephalopathy, seizures, and strokes [3]. A noninvasive and lower risk alternative to the Wada test would be preferred as it could increase patient willingness to undergo preoperative assessment. Resting-state functional magnetic resonance imaging (rs-fMRI) is a promising alternative [4]. This study will assess the concordance of memory lateralization results between rs-fMRI and the Wada test in pilot epileptic patients to determine the viability of rs-fMRI as an alternative to the Wada test.

Materials and Methods: rs-fMRI data was acquired in a pilot group of three patients and then analyzed using scripts based on the Analysis of Functional NeuroImages (AFNI) software suite [5]. Data was processed by first aligning and deobliquing. Then, the data was bandpass filtered from 0.005 to 0.1 Hz and volumes were blurred to 6 mm full width at half maximum before regression analysis was performed. The hippocampus and parahippocampal gyrus were the regions of interest (ROIs) and the retrosplenial cortex was the seed. The ROIs and seed were chosen to assess memory lateralization based on their role in the memory network [6]. Correlation maps for the ROIs based on the seed were created and a script was run to obtain the Z-score representing connectivity. rs-fMRI memory lateralization was determined by setting a threshold for correlation Z-scores of 0.5 and determining the brain hemisphere with the greater number of voxels above the threshold. The Wada test was performed by introducing propofol, an anesthetic, into one hemisphere of a patient’s brain followed by the administration of language and memory tests before repeating the process with the other hemisphere. The memory task score of each brain hemisphere is compared to determine Wada test memory lateralization. The rs-fMRI results were compared to the Wada test results to determine concordance in memory lateralization between the two modalities.

Results: Our analysis have yielded functional connectivity between retrosplenial cortex and hippocampus and parahippocampal gyrus. Comparing the memory lateralization results of the rs-fMRI and Wada test yielded concordant results in two patients with left lateralization (discrepancy of 0.5), and discordant results in one patient with discrepancy of 6.5, which does not reveal a conclusive relationship between the two.

Discussion: The inconclusive results may be due to a small sample size. Another possible explanation for the inconclusive results is interference from impaired neurovascular coupling (NVC). NVC is the mechanism by which transient neural activity evokes hemodynamic oscillations. Impaired NVC has been implicated in epilepsy and can lead to abnormal fMRI hemodynamics which would yield incorrect memory lateralization results [4].

Conclusion: This study has analyzed rs-fMRI and Wada test data from pilot epileptic patients and compared the memory lateralization results of the two modalities. The initial results of this study are inconclusive but demonstrate the feasibility of the methodology used for comparing the rs-fMRI and Wada test memory lateralization results. Further data acquisition through recruitment of additional epileptic patients should provide enough data needed to establish a proper conclusion. Interference from impaired NVC can be addressed by assessing the NVC of patients and ruling out the results of patients with impaired NVC, for example via the concurrent imaging of electroencephalogram and functional near infrared spectroscopy. For patients with normal NVC, we expect that rs-fMRI memory lateralization results may be concordant with Wada test memory lateralization results.

References:
[1] Sheng, J. Curr. Neuropharmacol. 2018. 16. 17-28.
[2] Englot, D. J. Neurology. 2012. 78. 1200-1206.
[3] Loddenkemper, T. Epilepsy & Behavior. 2008. 13. 551-553.
[4] Rolinski, R. Human Brain Mapping. 2020. 41. 3133-3146.
[5] Yuan, H. Brain Connectivity. 2014. 4, 690-701.
[6] Train, D. K. J Neurosurgery Pediatrics. 2022. 30. 394-399.