1501
Altered Amplitude of Low Frequency Fluctuation in Language Eloquent Areas in Patients with Medically-refractory Temporal Lobe Epilepsy
Li Jiang1,2, Stephanie Chen3, Lorenna Vidal4, Jiachen Zhuo1,2, Rao Gullapalli1,2, and Prashant Raghavan2
1Center for Advanced Imaging Research, University of Maryland Baltimore, Baltimore, MD, United States, 2Department of Diagnostic Radiology & Nuclear Medicine, University of Maryland Baltimore, Baltimore, MD, United States, 3Department of Neurology, University of Maryland Baltimore, Baltimore, MD, United States, 4Department of Radiology, Children's Hospital of Philadelphia, Philadelphia, PA, United States
1Center for Advanced Imaging Research, University of Maryland Baltimore, Baltimore, MD, United States, 2Department of Diagnostic Radiology & Nuclear Medicine, University of Maryland Baltimore, Baltimore, MD, United States, 3Department of Neurology, University of Maryland Baltimore, Baltimore, MD, United States, 4Department of Radiology, Children's Hospital of Philadelphia, Philadelphia, PA, United States
Synopsis
Temporal lobe epilepsy (TLE) is the most common type of epilepsy in adults. Language impairment can result from both continued seizures and surgical attempts to treat it. Thus, accurate preoperative assessment of language function is essential. Here we used resting-state fMRI to investigate the altered amplitude of low frequency fluctuation (ALFF) in language eloquent areas in medically-refractory TLE patients and its relationship with clinical language test measures. Our findings suggest that left TLE disrupts language function more than right TLE and that intrinsic spontaneous brain activity is altered even in the absence of detectable clinical language impairment.
Purpose
Temporal lobe epilepsy (TLE) is the most common type of epilepsy in adults [1, 2]. Surgery is the treatment of choice for medically-refractory TLE patients. Language impairment can result from both continued seizures and surgical attempts to treat it. Thus, accurate preoperative assessment of language function is essential. Task-based fMRI is the most commonly used noninvasive alternative to the intracarotid amobarbital test for assessing language function [3], however, its clinical utility and reliability may be limited by age, language barriers, attentiveness, or states of altered consciousness due to seizures and antiepileptic medications. In these situations, task-free approach such as resting-state fMRI (rs-fMRI) may be more reliable [4]. The amplitude of low frequency fluctuation (ALFF) derived from the rs-fMRI BOLD signal has been suggested to reflect the intensity of regional spontaneous brain activity [5]. In this study, we investigated the altered ALFF in language eloquent areas in medically-refractory TLE patients compared with healthy controls (HC), and examined the relationship between the ALFF and clinical language test measures.Materials and Methods
A cohort of 17 unilateral medically-refractory TLE patients (5 males/12 females, age of 36.1±13.24, range 21-64 years) undergoing presurgical evaluation were retrospectively reviewed from the Maryland Epilepsy Center between 2014 and 2019. The TLE patients were divided into left TLE (n=9) and right TLE (n=8) based on the seizure lateralization. All the patients underwent standard clinical preoperative evaluation and rs-fMRI scans. The comprehensive language tests included Controlled Oral Word Association test (COWA), Multilingual Aphasia Examination (MAE), and Boston Diagnostic Aphasia Examination (BDAE). Fifteen of 17 patients were left-hemisphere dominant and two patients had mixed language dominance. (Table 1).The rs-fMRI data were acquired either at a 3T Siemens Tim-Trio (n = 14) or Prisma Scanner (n = 3). The imaging parameters at the Tim-Trio scanner were TR/TE=2000ms/30ms, slice thickness=4mm, and resolution=3.75x3.75mm2 for 171 volumes, while those at the Prisma Scanner were TR/TE= 1600ms/30ms, slice thickness=3mm, resolution=3x3 mm2 for 250 volumes. Seventeen age-gender matched HCs (7 males/10 females, age of 36.70±13.24, range 20-64 years) were also retrieved and scanned on the Tim-Trio system.
All the rs-fMRI data were trimmed to 171 volumes across time and were preprocessed using the CONN toolbox (v18), especially, spatial smoothing with 8 mm FWHM), and temporally band-pass filtered (0.01~0.08 Hz). The ALFF analysis were performed utilizing the REST toolkit. ALFF map were obtained as the power spectrum of the Fourier Transform of the BOLD time series at each voxel. ALFF map was normalized by the global mean of ALFF in the brain for standardization (mALFF). The GLM with age corrected was performed to compare the mALFF group difference in the language network including the Broca’s area: the inferior frontal gyrus pars opercularis and pars triangularis (IFG-oper and IFG-tri) and the Wernicke’s area: Superior temporal gyrus (STG), middle temporal gyrus (MTG), supramarginal gyrus (SMG), and angular gyrus (AG) at both dominant and non-dominant language hemispheres. Significant difference was set as voxel-wise p < 0.008, Alphasim multiple comparison corrected p < 0.05 and cluster-size > 73 voxels.
Partial correlation analysis with age as covariate was performed to examine the relationship between regional mALFF values and clinical language testing scores (verbal fluency score and comprehension score) within the TLE group with significant level of p < 0.05.
Results
As shown in Table 2, the verbal fluency/COWA testing was utilized as a surrogate for Broca’s language function and 3 out of 15 patients were below average. Aural and Reading Comprehension was utilized as a surrogate for Wernicke’s language function and 5 out of 15 patients were below average.As shown in Figure 1 and Table 3, compared with HC group, the TLE group had significantly decreased mALFF in right SMG, right posterior STG, and right AG. The left TLE group showed significantly decreased mALFF in left and right SMG; and the right TLE group showed no difference. Compared with right TLE group, left TLE group had significantly decreased mALFF in left AG.
Significant positive correlation was observed between the averaged mALFF in right posterior STG and clinical comprehension language score (r = 0.465, p = 0.047; Figure 2). No other significant correlations were seen between mALFF at other language regions and clinical language testing scores. We also did not observe significant correlations between mALFF in the language eloquent areas and duration of epilepsy.
Discussion and Conclusion
In this study, TLE patients showed decreased mALFF in the right STG and right AG compared to HCs. Similar findings were reported by Singh et al (2020) where they observed decreased ALFF in the inferior parietal lobule and superior temporal gyrus in TLE patients [6]. We also report a greater decrease in mALFF in the left AG in left TLE patients compared to right TLE patients which suggests that left TLE disrupts language function more than right TLE. Noting that most of our patients showed average language level on the clinical language test, our findings may indicate that the intrinsic spontaneous brain activity is altered even in the absence of detectable clinical language impairment. The main limitations of our study are our modest sample size and lack of consideration of seizure frequency and interictal discharges.Acknowledgements
No acknowledgement found.References
1. Bell, B., et al., The neurobiology of cognitive disorders in temporal lobe epilepsy. Nat Rev Neurol, 2011. 7(3): p. 154-64.
2. Tellez-Zenteno, J.F. and L. Hernandez-Ronquillo, A review of the epidemiology of temporal lobe epilepsy. Epilepsy Res Treat, 2012. 2012: p. 630853.
3. Bauer, P.R., et al., Can fMRI safely replace the Wada test for preoperative assessment of language lateralisation? A meta-analysis and systematic review. J Neurol Neurosurg Psychiatry, 2014. 85(5): p. 581-8.
4. Smitha, K.A., et al., Resting fMRI as an alternative for task-based fMRI for language lateralization in temporal lobe epilepsy patients: a study using independent component analysis. Neuroradiology, 2019. 61(7): p. 803-810.
5. Zou, Q.H., et al., An improved approach to detection of amplitude of low-frequency fluctuation (ALFF) for resting-state fMRI: fractional ALFF. J Neurosci Methods, 2008. 172(1): p. 137-41.
6. Singh, T.B., et al., The Assessment of Brain Functional Changes in the Temporal Lobe Epilepsy Patient with Cognitive Impairment by Resting-state Functional Magnetic Resonance Imaging. J Clin Imaging Sci, 2020. 10: p. 50.
Figures
Table 1: Clinical Demographics of the medically-refractory patients (n
= 17).
Table 2: Clinical Language Testing Scores
Table 3. Group comparison of the mALFF
within language
eloquent areas
Figure
1: The mALFF difference within language eloquent areas among
groups (A) all TLE
patients and HCs. (B) left TLE and HCs; (C) left TLE
and right TLE patients. The significant difference was defined as
voxel-wise uncorrected p < 0.008, cluster-wise corrected p < 0.05 and
cluster-size > 73 voxels. L:
left hemisphere; R: right hemisphere. Hot color refers to increased mALFF and
cold color refers to decreased mALFF.
The Pearson
correlation between the language test scores and the mALFF in
the sub-regions within the language eloquent areas. (A) Significant clusters by
comparing all TLE patients and healthy controls. The regions of right superior
temporal gyrus (pSTG) was circled out. (B) Correlation
between the composite percentile scores and mALFF in the
right pSTG.