Late-Onset Temporal Lobe Epilepsy—Differentiating the Gray Matter from the Matter of Gray

Abstract

Considering the growing age of the world population, the incidence of epilepsy in older adults is expected to increase significantly. It has been suggested that late-onset temporal lobe epilepsy (LO-TLE) may be neurodegenerative in origin and overlap with Alzheimer's disease (AD). Herein, we aimed to characterize the pattern of cortical atrophy and cerebrospinal fluid (CSF) biomarkers of AD (total and phosphorylated tau, and β-amyloid) in a selected population of LO-TLE of unknown origin. We prospectively enrolled individuals with temporal lobe epilepsy onset after the age of 50 and no cognitive impairment. They underwent a structural MRI scan and CSF biomarkers measurement. Imaging and biomarkers data were compared to 3 retrospectively collected groups: (i) age- and sex-matched healthy controls, (ii) patients with mild cognitive impairment (MCI) and abnormal CSF AD biomarkers (MCI-AD), and (iii) patients with MCI and normal CSF AD biomarkers (MCI-noAD). From a pool of 52 patients, 20 consecutive eligible LO-TLE patients with a mean disease duration of 1.8 years were recruited. As control populations, 25 patients with MCI-AD, 25 patients with MCI-noAD, and 25 healthy controls were enrolled. CSF biomarkers returned normal values in LO-TLE, significantly different from patients with MCI due to AD. There were no differences in cortico-subcortical atrophy between epilepsy patients and healthy controls, while patients with MCI demonstrated widespread injuries of cortico-subcortical structures. Individuals with a late-onset form of temporal lobe epilepsy, characterized by short disease duration and normal CSF β-amyloid and tau protein levels, showed patterns of cortical thickness and subcortical volumes not significantly different from healthy controls, but highly different from patients with MCI, either due to AD or not.

As clinicians caring for older adults, we are often confronted with the intricate interplay between epilepsy and neurodegenerative disorders. With the global population aging, the incidence of late-onset epilepsy (LOE) is rising, leading us to ask: is epilepsy in older adults the final chapter of the book of epilepsy, or is it the first volume in neurodegeneration atheneum, or does it have a distinct identity? ¹

LOE has an unknown etiology in 20% to 30% of the cases, and it has been postulated that seizures might be the first symptom of impending dementia, especially Alzheimer's disease (AD). ² The study “Late-onset temporal lobe epilepsy: insights from brain atrophy and Alzheimer's disease biomarkers” by Ballerini et al provides valuable clues, particularly concerning late-onset temporal lobe epilepsy (LO-TLE). ³ Through rigorous examination of cortical and subcortical (hippocampus and amygdala) atrophy and cerebrospinal fluid (CSF) biomarkers, the study offers a clearer lens through which to view LO-TLE, potentially distinguishing it from AD.

The study enrolled 20 patients with LO-TLE (onset > 50 years), examining them alongside age- and sex-matched healthy controls and 2 groups with mild cognitive impairment (MCI), one with abnormal CSF AD biomarkers (MCI-AD) and another with normal CSF AD biomarkers (MCI-noAD). ³ Most of the patients were classified as mesial TLE and had normal MRI brain. The results are illuminating. Contrary to expectations that neurodegenerative markers might pervade in an epilepsy-prone older population, LO-TLE patients exhibited no significant cortical or subcortical atrophy when compared to healthy controls. These findings were in stark contrast to both MCI groups, which showed widespread and expected atrophy consistent with the underlying pathology. Moreover, CSF biomarkers in LO-TLE patients, including Amyloidβ (Aβ₁₋₄₂, Aβ₁₋₄₂/Aβ₁₋₄₀ ratio) and tau (total tau {tTAU}, phosphorylated tau {pTAU₁₈₁})proteins were similar to controls and MCI-noAD but significantly different than MCI-AD. From a clinical perspective, these findings are critical. They suggest that LO-TLE might not share the same pathological pathway as AD, as has been contemplated in other studies.

The study's most puzzling observation is the normal pTAU₁₈₁ levels in LO-TLE patients. This finding contrasts with data from younger TLE patients undergoing surgical resection, where elevated pTau is commonly observed. ⁴ A plausible explanation could be the study cohort's relatively short disease duration (mean of 1.8 years). Chronicity of epilepsy might be a more significant driver of pTau accumulation, pointing to the need for longitudinal studies to assess the evolution of tau pathology over time. Alternatively, it may hint at a distinct, non-amyloidogenic, non-tauopathic neurogenerative condition specific to chronic epilepsy, separate from the classical AD pathway. Indeed, in a review by Hickman et al, Late Onset Epilepsy of Unknown Etiology (LOEU) was classified into 3 types based on CSF amyloid (A) and tau (T) biomarkers, cognitive testing and FDG PET: (i) A⁺T⁺ suggesting epileptic preclinical or prodromal AD, (ii) A⁺T⁻ suggesting late-onset Aβ-related epilepsy, and (iii) A⁻T⁻ suggesting LOEU without positive AD biomarkers. ⁵ The study by Ballerini et al potentially suggests adding neuroimaging biomarkers, specifically cortical and subcortical atrophy into the mix. ³ While most of the conversation around LOEU has involved AD, it is important to note that other dementias are also associated with epilepsy. ⁶

From a clinical management standpoint, the study's findings could have several implications. They emphasize the necessity of differentiating between LOE with new-onset cognitive changes and neurodegenerative conditions in older adults, as these patients could benefit from different diagnostic approaches. The absence of pronounced atrophy and normal CSF biomarkers in LO-TLE might justify a more optimistic prognosis, potentially influencing discussions about life planning and management of comorbidities. This study indirectly supports advanced neuroimaging techniques and CSF biomarker analyses in routine clinical practice of LOE of unknown etiology, where feasible, to enhance diagnostic precision.

LO-TLE did not differ from MCI groups in years of disease duration and/or onset age. Previous studies have shown that with LOE and cognitive changes, both might start simultaneously in relatively younger patients and that these patients might have an accelerated disease course.^7,8 This finding could suggest that cortical atrophy in LO-TLE might become more prominent over time, eventually “catching up” with patterns seen in MCI-AD and MCI-noAD. Thus, regular follow-ups and a high index of suspicion remain essential in this patient group.

An exciting avenue for future research involves using amyloid PET imaging, which is now available clinically in the United States. ⁹ The ability to directly visualize amyloid deposition could provide critical insights into the neuropathological processes at play in LO-TLE. Additionally, emerging plasma biomarkers, such as pTau₂₁₇, hold promise for non-invasive monitoring of neurodegenerative processes. ¹⁰ Suppose these new diagnostic tools can effectively distinguish between LO-TLE and early-stage AD. In that case, they may codify our approach to managing epilepsy in older adults and clarify the links between chronic epilepsy and neurodegeneration.

Additionally, longitudinal studies should track CSF biomarker changes and cortical atrophy progression in LO-TLE patients. Exploring the exact mechanisms by which chronic seizures might induce or accelerate tau pathology will provide critical insights. Further research into whether distinct epileptic syndromes have unique biomarker profiles could lead to more precise categorizations and diagnostic strategies.

The relatively benign seizure outcomes observed in the LO-TLE cohort, with 75% achieving seizure remission, also warrant further investigation. Understanding why older patients respond more favorably to anti-seizure medications than their younger counterparts might reveal age-related differences in epileptogenesis and drug metabolism.

In conclusion, this study provides a unique picture of LO-TLE, distinguishing it from the neurodegenerative processes underlying AD. While these findings may not yet be definitive enough to change clinical practice entirely, they emphasize the importance of a nuanced approach to diagnosing and managing epilepsy in older adults. By focusing on the distinct biomarker and imaging profiles, clinicians can better tailor diagnostic strategies, improve patient outcomes, and anticipate potential neurodegenerative transitions.

After all, in the complex landscape of aging brains, it's essential to distinguish the forest from the trees—or, more precisely, the tau from the TLE.