Authors :
Presenting Author: Rani Sarkis, MD, MSc. – Brigham and Women's Hospital, Harvard Medical School
Sara Lariviere, PhD – Postdoctoral Fellow, Neurology, Brigham and Women's Hospital; Frederic Schaper, MD, PhD – Brigham and Women's Hospital; Page Pennell, MD – Neurology – University of Pittsburgh Medical Center; Steven Stufflebeam, MD – Radiology – Massachusetts General Hospital; Geoffrey Young, MD – Radiology – Brigham and Women's Hospital; Michael Fox, MD, PhD – Neurology – Brigham and Women's Hospital; Gad Marshall, MD – Neurology – Brigham and Women's Hospital
Rationale:
Epilepsy is common after the age of 60 with around half of new-onset patients showing no identifiable cause. This condition is termed “late onset unexplained epilepsy” (LOUE). While the imaging profiles of individuals with LOUE are currently lacking, the prevailing hypothesis is that LOUE localizes predominantly to the medial temporal lobe and may be tied to temporal lobe epilepsy (TLE) or to an underlying neurodegenerative condition such as Alzheimer’s disease (AD). To test this hypothesis, we mapped cortical and hippocampal atrophy in LOUE and assessed whether atrophy profiles are similar to younger onset TLE or to AD.
Methods:
We studied 47 individuals with LOUE (age: 69.6±5.9 years) from Brigham and Women’s Hospital and 48 age-, sex-, and education-matched healthy controls (age: 70.6±5.2 years) from the Harvard Aging Brain Study. From T1-weighted MRIs, we measured cortical and hippocampal thickness in every participant. Linear models compared atrophy in participants with LOUE to controls, correcting for age, sex, and intracranial volume, as well as multiple comparisons using the false discovery rate (
pFDR). Subgroup analyses further compared differences in cortical and hippocampal atrophy between participants with LOUE (
i) with or without an APOE ε4 allele and (
ii) with a Clinical Dementia Rating Global score of 0.5 or 0.
To assess the similarity of atrophy profiles in LOUE to younger onset TLE and AD, we repeated the above linear models in large, multi-site cohorts (Ke et al.), comparing: (
i) TLE (31.3±10.9 years) versus age matched controls (
nTLE/HC=94/120), and (
ii) AD (75.0±8.1 years) versus age-matched controls (
nAD/HC=79/117; from the Alzheimer’s Disease Neuroimaging Initiative). We then spatially correlated patterns of cortical and hippocampal atrophy in LOUE to those observed in TLE and AD. The statistical significance of these spatial correlations was assessed using spin permutation tests.
Results:
Compared to controls, participants with LOUE showed significant cortical atrophy in bilateral entorhinal cortex and calcarine cortex, as well as in left postcentral gyrus (all
pFDR< 0.05;
Fig. 1A). No significant differences were observed when directly comparing subgroups of LOUE, although participants with an APOE ε4 allele, as well as those with a Clinical Dementia Rating of 0.5, displayed slightly more widespread cortical atrophy notably in the fronto-parietal cortices and more prominent hippocampal atrophy (
puncorr< 0.01). The pattern of cortical atrophy across all participants with LOUE was spatially similar to that of younger onset TLE (
r=0.16,
pspin=0.034;
Fig. 1B), whereas the pattern of hippocampal atrophy in LOUE was more similar to that observed in AD (
r=0.14,
pspin=0.0058;
Fig. 1C).
Conclusions:
Group-level atrophy patterns in LOUE are distributed across medial temporal lobe regions and the postcentral gyrus, and reflect, to some extent, known cortical patterns of atrophy in younger TLE and hippocampal atrophy in AD. Future analyses will address where each individual LOUE subject lies along the spectrum of AD and younger onset TLE.
Funding:
American Epilepsy Society Junior Investigator Award, NIH K23NS119798 (RAS), Canadian Institutes of Health Research (SL).