Victoria L. Morgan1, Ahmet Cakir2, Benjamin N. Conrad1, Bassel Abou-Khalil3, Adam W. Anderson1,4, Zhaohua Ding1, and Bennett A. Landman1,2
1Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, 2Electrical Engineering and Computer Science, Vanderbilt University, Nashville, TN, United States, 3Neurology, Vanderbilt University, Nashville, TN, United States, 4Biomedical Engineering, Vanderbilt University, Nashville, TN, United States
Synopsis
Temporal lobe epilepsy (TLE) is a common and relatively
homogeneous form of epilepsy in which seizures originate in the mesial temporal
regions including the hippocampus and propagate across the brain. This work
represents the first step in characterizing the functional (FC) structural (SC)
network connectivity evolution in TLE using MRI. We found consistent decreases
in ipsilateral hippocampus and insula FC and SC primarily after 10 years of
duration of disease in patients with seizure freedom after surgery. In those
with seizure recurrence, there were more severe bilateral hippocampal SC
decreases when compared to those with seizure freedom.Introduction
Temporal lobe epilepsy (TLE) is a common and relatively
homogeneous form of epilepsy in which seizures originate in the mesial temporal
regions including the hippocampus. The seizures propagate across temporal lobe
networks and throughout the brain. When refractory to pharmacological therapy,
seizures can occur in this manner over several years or decades over the
duration of disease. Our working model consists of a temporal and cingulate
network which evolves over years in both functional and structural
connectivity. This work represents the first step in characterizing this
network evolution using MRI.
Methods
We have currently recruited 26 TLE patients and 26 age (+/-
2 years) and gender matched controls. This cross sectional analysis included 20
TLE patients, 15 who were seizure free after resection of the temporal lobe or
anterior temporal lobe (sz free group, 7F, 10 right TLE, duration 17.1 +/- 13.1
yrs) and 5 patients who had seizure recurrence (sz recur group, 4F, 5 right
TLE, duration 18.4 +/- 16.3 yrs). Subjects were scanned on a 3T MRI scanner
using resting state fMRI for functional connectivity (FC) (TR=2 s, 3x3x4 mm3, 300 vols) and diffusion weighted MRI for structural connectivity (SC) (b=1600 s/mm2,
92 directions, 2.5 x. 2.5 x 2.5 mm3, 3 averages).
The nodes of interest were defined ipsilateral and contralateral
to seizure onset (hippocampus: HIPI, HIPC; insula: INSI, INSC; thalamus: THALI,
THALC). In addition, two midline bilateral structures were defined (precuneus:
PREC; midcingulate cortex: CING). For FC, the preprocessed time series were
averaged across each node and pairwise correlations were determine and
transformed to Z scores. For SC, the FSL probabilistic fiber tracking protocol
was computed separately with each region as seed with the other seven regions
as targets and terminations. The SC was computed as the percent of total fibers
originating from the seed reaching the target. All patient FC and SC values were
transformed to the percent change from age matched control. The sz free group
was divided into those with less than 10 years (n=5) and greater than or equal
to 10 years (n=10) disease duration. T-tests were used to determine difference
from controls and between sz free and sz recur groups (p<0.01).
Results
Figure 1 shows network diagrams with each node as a red
circle. Paths with decreased connectivity from controls are indicated by dashed
lines. There were no increases. Figure 2
shows the change from control for each sz free patient (FC blue, SC red) for
two paths (INSI-CING, HIPI-HIPC) shown in Figure 1. Zero on the y-axis represents
the age corrected value of the control subject. The 10 year line dividing the
data in the left and right columns of Figure 1 is indicated on the x-axis.
Figure 3 shows the path and the associated individual data of the path where
the SC of the sz free group was increased over the sz recur group. There were
no differences between groups in FC.
Discussion and Conclusions
These results suggest an evolution of both FC and SC
networks over years of duration in TLE. The decreases in FC in HIPI-HIPC and
HIPI-PREC at durations 10 years or greater are consistent with the literature
[1]. In addition, FC decreases from the INSI are also not surprising given the
known role of the insula in TLE seizures [2]. In SC, the decreases in the sz
free group were also generally consistent with other reports [3].
This work includes several novel concepts. The first is the
age control correction of the patient data. We enrolled an individual age
matched control for each patient. We then fit the controls vs. age and subtracted
this fit from the patient and divided by it to get a percent change from age
corrected control. Second, we grouped patients into sz free and sz recur groups
and detected SC differences in bilateral hippocampal paths. Third, we investigated
early and late changes in connectivity and found more significant changes later
in duration. Finally, out of 28 paths investigated, we found 2 consistent
ipsilateral paths with decreases in both FC and SC.
In summary, we found consistent decreases in ipsilateral
hippocampus and insula FC and SC primarily after 10 years of duration of
disease in TLE patients with seizure freedom after surgery. In those with
seizure recurrence, there were more severe bilateral hippocampal SC decreases
when compared to those with seizure freedom.
Acknowledgements
Funded by R01 NS75270 (VLM)References
[1] Periera FRS et
al. BMC Neuroscience 2010;11:66.
[2] Isnard J, et al. Annals of Neurology 2000; 48: 614-623.
[3] Besson P. et al., NeuroImage 2014; 100:135-144.