Does Alzheimer’s disease with mesial temporal lobe epilepsy represent a distinct disease subtype?

Abstract

Alzheimer’s disease(AD) patients have a high risk of developing mesial temporal lobe epilepsy(MTLE) and subclinical epileptiform activity.MTLE in AD worsens outcomes.Therefore, we need to understand the overlap between these disease processes.We hypothesize that AD with MTLE represents a distinct subtype of AD, with the interplay between tau and epileptiform activity at its core.We discuss shared pathological features including histopathology,an initial mesial temporal lobe(MTL) hyperexcitability followed by MTL dysfunction and involvement of same networks in memory(AD) and seizures(MTLE).We provide evidence that tau accumulation linearly increases neuronal hyperexcitability,neuronal hyper-excitability increases tau secretion, tau can provoke seizures and tau reduction protects against seizures.We speculate that AD genetic mutations increase tau, which causes proportionate neuronal loss and/or hyperexcitability, leading to seizures.We discuss that tau burden in MTL predicts cognitive deficits among 1)AD and 2)MTLE without AD. Finally, we explore the possibility that anti-seizure medications improve cognition by reducing neuronal hyper-excitability, which reduces seizures and tau accumulation and spread.

BACKGROUND

Dementia is the most costly disease in the US¹. Alzheimer’s disease (AD) is the most common type of dementia.AD is a public health crisis, the leading cause of disability, morbidity and mortality among older adults,¹. AD pathology is characterized by histopathological evidence of senile or beta-amyloid plaques (Aβ) and neurofibrillary tangles (NFTs)² characterized by accumulation of abnormally phosphorylated tau (pTau).

Epilepsy is the third most common neurological condition among adults over 65³ and is also considered a public health imperative. Its incidence is highest among the elderly age group³. People with epilepsy (PwE) have a high degree of associated comorbidity, especially in the elderly⁴.

AD patients have a substantially higher risk of developing epilepsy than the general population especially mesial temporal lobe epilepsy (MTLE) and late-onset PwE have a substantially higher risk of developing AD^5-8. Recent meta-analyses have confirmed a very high prevalence of epilepsy and epileptiform activity in AD^9,10. There is a large body of evidence that MTLE -related seizures are very common in AD patients, and they worsen prognosis and may even be associated with shorter life expectancy¹¹. Despite the high incidence, co-occurrence, and worse outcomes, we do not understand the clinical presentation, neuropathology, clinical course, and outcomes of patients with AD and co-morbid epilepsy. To understand these, the interaction between the two disease processes needs to be delineated.

Is AD with MTLE a distinct subtype of AD with pTau as the predominant driver behind these disease processes?

We propose that AD with MTLE represents a clinically distinct subtype of AD with distinct neuropathology, clinical features, and a more aggressive disease course with worse outcomes than AD patients without MTLE. Herein, we provide evidence that these two disease processes are characterized by excessive tau pathology, and speculate that tau may provide the link between these two diseases. Further, we explore the possibility that the interplay between tau, neuronal hyperexcitability and epileptiform activity is at the core of this distinct subtype of AD.

High co-occurrence and bi-directional association between MTLE and AD

Unprovoked seizures occur in 10-64% of AD patients at varying stages of the disease process¹². Patients with AD ≥ 65 years of age have an 8-to-10-fold higher rate of unprovoked seizures compared to the general population⁶. In addition, AD patients have a considerably high risk of developing epilepsy especially MTLE⁷. Early onset AD (<65 years at onset) has an even higher association with epilepsy compared to late onset AD¹³.

Older PwEhave a relative risk of 1.5 of developing dementia over a span of 8 years^8,14. Late-onset PwE have a substantially higher risk of developing dementia especially AD⁵. AD and APOε4 have been found to be among the leading risk factors for the development of late-onset epilepsy (epilepsy onset at or after the age of 65)¹⁵.

In addition to seizures and epilepsy, many AD patients show evidence of subclinical interictal epileptiform activity and/or electroencephalographic (EEG) seizures without a known history of epilepsy^16-19. As many as 42% of AD patients demonstrate subclinical interictal epileptiform discharges, especially of the mesial temporal regions^16-20. . In addition, silent hippocampal seizures occur in AD patients¹⁸. Up to 90% of epileptiform activity is seen with extended EEG recorded exclusively during sleep¹⁶.

These findings suggest that occult mesial temporal hyper-excitability may contribute to AD pathogenesis.

The possible shared pathology between late onset MTLE and AD:

In both animal and human studies, increasing evidence suggests overlap and shared pathological features between late-onset MTLE and AD. AD and MTLE have several intersecting neuroimaging and histopathological features²¹. Both diseases primarily affect the mesial temporal lobe and memory^16-18. Amnestic MCI and early AD patients demonstrate an initial mesial temporal lobe hyperexcitability followed by mesial temporal lobe dysfunction²². This is similar to mesial temporal lobe hyperexcitability observed during seizures in MTLE and the mesial temporal lobe dysfunction observed in between seizures.

There is also evidence that the same cortical networks that are involved in memory and altered in AD, are also affected in MTLE²³. Hippocampus and entorhinal cortex are both involved in memory storage and retrieval as well as in generating mesial temporal seizures²³. Alterations in local limbic circuits including papez circuit are often observed in both AD and MTLE ²³. Hippocampal coupling with association cortices has also been shown to be involved in memory consolidation especially long term memory, in both animal and human studies^24,25. Seizures in the mesial temporal lobe epilepsy also follow the same cortical functional connections of the entorhinal cortex and hippocampus.

Histopathological features of AD and MTLE:

Both disease processes demonstrate early involvement of the entorhinal cortex and hippocampus. AD associated NFTs are composed of abnormally phosphorylated tau protein. The pattern of NFT deposition in AD is relatively stereotyped and the degree, as well as areas of NFT involvement, tend to be associated with the degree of cognitive impairment and the disease duration^2,26. Senile,neuritic or beta-amyloid (Aβ) plaques represent another cardinal histopathological findings among AD patients ^2,27.

In AD, tau pathology involves entorhinal cortex early-on and loss of neurons in layer II and IV of entorhinal cortex is one of the earliest changes seen^28,29. Hippocampal CA1 (cornu ammonis 1) receives its inputs from the entorhinal cortex and is also involved next in the pre-clinical stages of disease²⁸. It spreads from cell to cell transneuronally and trans-synaptically secondary to neuronal activity. Tau deposits follow patterns of entorhinal functional connectivity. From the entorhinal cortex, tau spreads to involve limbic regions including hippocampus, parahippocampal gyrus, amgydala, perirhinal cortex and eventually to inferior temporal region and the basal forebrain^28,30. During the later stages of disease, neocortex especially precuneus and lateral parietal regions become involved^28,30.Other histopathological findings in AD tend to also occur in the hippocampus. These include granulovacuolar degeneration, hirano bodies and synaptic loss. ².

Studies have shown that drug resistant MTLE patient undergoing temporal lobectomy show findings similar to AD histopathology and involve similar regions. Most commonly stigmata of high pTau in the mesial temporal structures on immunohistochemistry is observed^31,32. Less commonly, amyloid plaques may be seen in MTLE patients undergoing temporal lobectomy³³. Other less common findings that may be seen in both conditions include reduction of calbintin-D28k (calcium binding protein) in dentate gyrus³⁴ and hippocampal sclerosis³⁵.

Revised BRAAK staging shows that, during the pretangle stage of AD, pTau accumulates in subcortical locus coeruleus before any cortical regions are involved, including mesial temporal regions³⁶. While locus coeruleus in autopsy specimens of MTLE patients has not been studied for evidence of pTau, it is interesting to note that neurostimulation of locus coeruleus as a target for the treatment of epilepsy has shown successful results in improvement of epilepsy^37-39. Therefore, potential role of locus coeruleus in MTLE needs to be elucidated.

Hippocampal neuronal hyperexcitability and dysfunction is at the core of both disease processes:

Seizures are defined as abnormal, synchronous excitation of neurons in the brain. It is well-known that network and neuronal hyperexitability is the primary cause of seizures and hippocampal neuronal hyperexcitability is seen mesial temporal lobe seizures . While during ictal states, hippocampal hyperexcitability is seen, interictally hippocampal dysfunction is observed. Very similar to this, recent evidence has suggested that AD also begins with network hyperexcitability. In earlier stages of AD,,in patients with subjective cognitive impairment or MCI and some cognitively healthy adults with familial risk for AD⁴⁰, hippocampal hyperactivation is observed⁴¹ just like the hippocampal hyperactivation seen in MTLE. Hippocampal hyperactivation is accompanied by cortical thinning in the regions of the brain classically implicated in AD⁴⁰. Hippocampal hyperexcitability is also evident in AD patients from the increased amount of subclinical IEDs and silent seizures reported on EEG of these patients^16-18.

It has been proposed that hippocampal activity follows an inverted U shape pattern and early hyperactivity is followed by hypoactivity in the later stages of AD²². Hippocampal hyperactivity seen in early stages of cognitive impairment can manifest as epileptiform discharges and seizures since ictal discharges commonly arise from the hippocampus in MTLE. Similarly, hippocampal hypoactivity⁴¹, seen in the later stages of cognitive impairment and overt AD is a manifestation of an increasingly diseased and dysfunctional hippocampus. This is similar to hippocampal dysfunction often seen in advanced MTLE.

Network dysfunction in AD has also been observed in functional magnetic resonance imaging (fMRI) studies of these patients during memory tasks and involves the regions typically implicated in MTLE. Network alterations are most prominent in mesial temporal lobe and neocortical regions involved in memory, including precuneus and posterior cingulate⁴².

Moreover, hippocampal atrophy and often visible sclerosis is a common neuro-imaging abnormality often observed in both conditions²¹ (Figure 1).

Figure 1: Similarities between neuro-imaging findings of AD and MTLE.

Note the atrophied hippocampi bilaterally, which may be sclerosed in either of the conditions.

This evidence of hippocampal hyper and hypoactivity at various stages of cognitive impairment and the presence of hippocampal sclerosis (Figure 1) can explain the increased co-occurrence of epileptiform activity and seizures in the mesial temporal regions in patients with AD.

Memory circuits in AD and seizure network in MTLE involve the same brain regions:

Papez circuit (Figure 2) connects hippocampus via fornix to mammillary body and anterior nucleus of thalamus which sends projections via cingulate gyrus to parahippocampal gyrus which connects back to the hippocampus, and plays a major role in memory consolidation²³. Various degrees of alteration in the papez circuit is observed in AD. Similarly, alteration of the papez circuit is often implicated in many PwE especially those with MTLE, and various regions of the papez circuit are increasingly being studied and used as therapeutic targets for neuro-modulation in the management of drug-resistant epilepsy²³.

Figure 2: Papez Circuit:

The memory pathway and the pathway for seizures are the same. The interaction between different structures is demonstrated. A: Shows a simplified version of the Papez circuit, while B demonstrates the structures on a sagittal MRI scan. ATN: Anterior thalamic nucleus.

The connection between the hippocampus and cortex is required for memory formation in the brain²⁵. Interictal epileptiform discharges (IEDs) in the hippocampus on depth recording have been shown to disrupt memory retrieval, consolidation and cognition in rat models²⁵. It has been hypothesized that IEDs provide a way by which abnormal activity is transmitted to cortical brain regions and networks involved in memory consolidation. Thus, IEDs not only disrupt hippocampal networks, but also cortical networks and areas connected to hippocampus²⁵. Depth electrode studies of MTLE PwEhave shown that hippocampal IEDs result in transient cognitive impairment and interrupt short-term memory retrieval and memory maintenance²⁴. In humans, episodic memory formation involves the connections between hippocampus, other mesial temporal lobe structure including entorhinal cortex and the association neocortex including mesial prefrontal cortex, anterior cingulate and posterior parietal cortex such as precuneus²⁴. Memory consolidation involves the transfer of memory from the hippocampus to these association cortices for long-term storage. The cortical-hippocampal network involved in memory consolidation includes the same regions that have been shown to synchronize in MTLE during the generation and progression of seizures on depth electrode recording²³.

The default mode network (DMN) includes precuneus, posterior cingulate, lateral and inferior parietal lobes, and temporal and medial frontal lobes (Figure 3). It is well-established that DMN is impaired in AD²². In addition, recent evidence has demonstrated impaired DMN in MTLE as well³³.

Figure 3: Default Mode Network:

The default mode network includes precuneus (B, yellow), posterior cingulate (B, yellow), lateral and inferior parietal lobes (A, blue), and temporal (A and B, green), and medial frontal lobes (A and B, mustard)

Thus it appears that both regional and cortical networks implicated in memory dysfunction in AD and in MTLE related seizures involve the same brain regions.

Tau pathology, network dysfunction, neuronal hyper-excitability and seizures:

Both animal and human studies have shown that AD pathology, especially pTau is associated with hippocampal-parahippocampal network dysfunction, aberrant neuronal firing, and neuronal hyperactivity²².

Animal studies of epilepsy and AD have shown that tau/A152T mutation, which confers a risk for AD and other taupathies can result in pTau accumulation, progressive neuronal loss in the hippocampi and excess tau aggregation which stimulates pre-synaptic glutamate release which in turn increases neuronal hyper-excitability⁴³. Tau also has a linear correlation with neuronal hyperexcitability and network dysfunction^34,44,45.

Neuronal activity and excitability regulate tau secretion and spread⁴⁶. In a study, human pluripotent stem cell neurons were treated with picrotoxin, a non-competitive blocker of GABA receptor chloride channel, to induce neuronal hyper-excitability. Picrotoxin-induced neuronal activity was found to be associated with both tau release and propagation in vitro⁴⁶.

Tau can provoke seizures^34,44,47. Mice models have demonstrated that total tau levels parallel with seizure severity, and in one study, mice with the highest tau levels had the most severe seizures⁴⁷. Thus epileptiform activity and seizures in AD can lead to further worsening and accumulation of Tau, worsening the disease process.

Several studies have demonstrated that reduction of tau reduces network hyper-excitability and is protective against seizures^34,44,47. Genetic ablation of tau has been shown to reduce hyperexcitability in AD mice⁴⁷. DeVos et al used anti-sense oligonucleotides to reduce endogenous tau expression and found that mice with less tau had less severe seizures than controls⁴⁷.

Similar evidence of the association between tau pathology and neuronal dysfunction and hyper-excitability exists in human studies. EEG and magnetoencephalography (MEG) have been extensively studied for network dysfunction among AD patients. It has been shown that slow background activity and reduced alpha synchrony on EEG and MEG correlate with pTau burden and cognitive decline among AD patients⁴⁸.

There are alternative hypotheses by which AD pathology leads to neuronal hyperexcitability and seizures. These include Aβ induced hyperexcitability, impaired glial transporters leading to accumulation of extrasynaptic glutamate, tyrisone kinase fyn over-expression, altered N-methyl-D-aspartate (NMDA) activity, dysfunction of GABAergic neurons especially in the hippocampus and enhanced synaptic transmission secondary to low amyloid beta among others¹⁹. The role of Aβ deposits in promoting seizures by altering neuronal membrane properties and destabilizing the synaptic cleft is well documented by animal models^49,50.

There is no controversy that Aβ, pTau, and tyrosine kinase Fyn all have a role to play in seizure generation in AD. However, several mice models have shown that synaptic and network dysfunction and toxicity included by Aβ and fyn is dependent on tau levels^34,44. The dendritic function of tau, has been implicated in mediating this toxicity⁴⁴. Studies have also shown that reduction in tau has been shown to prevent synaptic transmission, enhance inhibitory current, reduce excitation/inhibition imbalance, reduce NMDA receptor mediated dysfunction, prevent cognitive deficit and prevent spontaneous epileptiform discharges that maybe induced by fyn and Aβ³⁴. In mice with over-expression of fyn and Aβ , decreasing tau reduces the network hyper-excitability and severity of seizures³⁴.

Therefore, the constellation of human and animal studies provides evidence that is overall most compelling for the correlation between excess tau, neuronal hyper-excitability, dysfunction, and seizures. Based on the evidence detailed above, we hypothesize that in AD patients with MTLE, tau leads to neuronal excitability and provokes epileptiform activity and seizures, and neuronal hyperexcitability, epileptiform activity and seizures increase tau accumulation and spread, thereby creating a vicious cycle (Figure 4A). This vicious cycle accelerates the progression of disease, leading to worse cognitive and clinical outcomes. The possible mechanisms by which tau causes hyperexcitability in AD and MTLEs are summarized in Table 1. We speculate that these mechanisms are the same in both pathologies. Likewise, we speculate that tau targeting therapies reduce tau build-up and propagation, which in turn reduce neuronal hyperexcitability, epileptiform activity and seizures, and anti-seizure medications reduce neuronal hyper-excitability and seizures and, as a consequence, accumulation, and spread of tau (Figure 4B).

Figure 4: The interplay between neuronal hyper-excitability, epileptiform activity, seizures and tau in Alzheimer’s disease and mesial temporal lobe epilepsy.

A: Tau accumulation leads to neuronal excitability which manifests as epileptiform activity and seizures. This neuronal hyperexcitability, epileptiform activity and seizures in turn increase tau accumulation and spread thereby creating a vicious cycle resulting in accelerating progression of disease, cognitive decline and worse outcomes. B: Tau targeting therapies reduce tau buildup and propagation which in turn reduces neuronal hyperexcitability and seizures. Anti-seizure medications reduce neuronal hyper-excitability, epileptiform activity and seizures which in turn reduce accumulation and spread of tau.

Table 1:

Possible Mechanisms by which tau leads to neuronal hyperexcitability in AD and MTLE^19,34,44,70

The dendritic function of tau mediates toxicity induced by Aβ and tyrosine kinase Fyn

Tau is associated with the depletion of Kv4.2, a dendritic potassium channel which in turn leads to dendritic hyperexcitability

Tau facilitates the enhancement of Aβ-induced defects in axonal transport.

Excess tau releases pre-synaptic glutamate.

Tau leads to enhanced NMDA receptor-mediated dysfunction

Tau is associated with increased excitation/inhibition imbalance

Tau leads to increased synaptic transmission

Tau reduces inhibitory current,

AD: Alzheimer’s disease, MTLE: Mesial temporal lobe epilepsy, Aβ: Amyloid beta

Genetic mutations in AD, increased tau burden, and increased epileptogenicity:

Apolipoprotein ε4 (APOε4) allele is a well-established risk factor for developing AD and is associated with early age of disease onset. Increased occurrence of spontaneous seizures has been reported among mice models of APOε4⁵¹. More recently, APOε4 has been identified as a risk factors of late onset epilepsy¹⁵. This has been hypothesized to be secondary to decreased CNS inhibitory tone and/or increased excitatory tone leading to increased neuronal network hyperexcitability⁵¹. In mice models of AD, APOε4 has been shown to contribute to the tau-dependent reduction in GABAergic interneurons in the dentate with resultant hippocampal hyperexictability⁵². In mice models, APOε4 has also been shown to worsen tau-mediated neurodegeneration significantly ⁵³. APOε4 mice have significantly higher levels of tau and a significantly greater degree of somatodendritic tau redistribution,resulting in more significant brain atrophy and neuro-inflammation⁵³. These findings suggest that APOε4 leads to higher tau levels in the AD brain. We hypothesize this higher accumulation of tau, in turn, may contribute to seizures.

Mutations in amyloid precursor protein (APP), including in Down’s syndrome patients, and in presenilin I (PSEN1) and presenilin 2 (PSEN2) are not only linked to the familial autosomal dominant form of AD but also confer a high risk of myoclonus and seizures⁵⁴. The underlying neuropathology contributing to seizures in PSEN mutations is yet to be determined, but some cases have been reported of underlying pathological changes of ectopic neurons with NFTs and tau deposits in white matter as well as hippocampal sclerosis⁵⁵. There, it is likely that even in PSEN mutation, alterations in tau level and NFTs may play a role in seizure occurrence.

Overall, these findings support the possibility that genetic mutations in AD lead to an increased tau burden, which may lead to seizures.

Tau pathology is the predominant predictor of cognitive deficits among AD patients:

The best predictors of global cognitive scores on AD neuropathological studies are total neurofibrillary tangle count in area 9 in the frontal lobe and entorhinal cortex, and the neuronal count in CA1 of the hippocampus⁵⁶. In addition, the regional burden of NFT strongly correlates with clinical manifestation and region/domain-specific cognitive scores among AD patients⁵⁷. NFTs appear to parallel both the disease duration and severity of AD²⁶. In contrast, amyloid volume is not significantly associated with the disease duration, degree of cognitive decline, or disease severity⁵⁶.

Tau PET using tracer ¹⁸F-AV1451 shows the regional distribution of tau within the human brain. Recent studies have shown that tau pathology, as demonstrated by Tau PET, has a robust correlation with anatomical, clinical manifestation, and cognitive decline among AD patients^58,59.

Thus it appears that pTau burden is the predominant predictor of global and region-specific cognitive deficits among AD patients.

Tau pathology is the predominant predictor of cognitive deficits among MTLE patients WITHOUT AD:

In a study of 138 deceased patients with drug-resistant epilepsy, postmortem Braak staging was done. The study found a correlation between epilepsy-related trauma to have an association with tau accumulation. Also, asymmetric patterns of tau deposits were noted within the sclerosed hippocampus of these PwE³².

Another study of 33 drug-resistant MTLE patients with hippocampal sclerosis who underwent temporal lobectomy after age 50 demonstrated that 31 (94%) patients had evidence of abnormally phosphorylated tau pathology on histopathological analysis. In addition, the degree of tau pathology correlated with the degree of verbal learning and naming scores one year after the surgery, and the patient with the highest tau burden developed AD 9 years after the resection³¹ . A significant association between a history of secondary generalized seizures and tau burden was also found³¹.

Thus, it appears that, tau pathology burden in the mesial temporal structures is the predominant predictor of cognitive deficits among PwE^31,32.

Does tau burden cause neuronal loss, which in turn causes seizures in AD?

Neuropathology analysis has shown that tau burden correlates with neuronal loss in AD patients²⁸. Neurofibrillary tangles have been shown to have an inverse relation with the neuronal number in AD²⁸. In addition, there is evidence to suggest that AD patients with generalized motor seizures have more significant brain atrophy and neuronal loss than AD patients without seizures, specifically in large pyramidal cells in the parietal cortex (area 7) and parahippocampal gyrus (area 28)^60,61.

These findings raise the possibility that tau pathology leads to neuronal loss, which in turn causes seizures in AD patients.

MTLE is associated with faster cognitive and clinical decline among AD patients:

Both animal and human studies have shown that epileptiform activity and seizures, especially those involving the mesial temporal lobe, can impact memory and cognition adversely in both acute and chronic stages of AD.

While epileptiform abnormalities may be a consequence of neurodegeneration, there is evidence to suggest that their presence is associated with faster cognitive impairment. Epileptiform activity results in temporary cognitive impairment and interrupts with short-term memory formation and retrieval in intracranial studies of mesial temporal regions in both animals and humans^23,25. In addition to acute effects, epileptiform discharges and seizures can have a chronic impact on cognition as well. Hippocampal network remodeling is a long-term sequela of epileptiform activity, which can adversely affect memory and cognition¹⁹.

In a case-control study, AD patients with seizures have been shown to have a more rapid language and cognitive decline over a 12-month period compared to AD patients without seizures⁶². A recent prospective study showed that AD patients with subclinical epileptiform activity on 24 hour video EEG without overt clinical seizures also experienced a faster decline in global cognition when compared to AD patients without subclinical epileptiform activity¹⁶. Both executive function and mini-mental state examination (MMSE) were found to decline faster among these patients. In a study of Down syndrome patients with AD, patients who experienced seizures were found to have more profound cognitive decline than those without seizures¹⁹.

Recent evidence also points towards the neurodegenerative effect of long-lasting epilepsies on the cortex. Recent longitudinal studies have shown that patients with uncontrolled focal epilepsy can experience widespread cortical thinning, especially in the frontal and temporal lobes. This effect was reduced for patients who underwent epilepsy surgery to control seizures^63,64.

Overall, these findings suggest that subclinical epileptiform activity and clinical seizures especially in the temporal region, are associated with accelerated cognitive decline among AD patients.

Anti-seizure medication reduce neuronal hyperexcitability and seizures, and can potentially slow cognitive decline among AD patients:

It has been shown that N-methyl-D-aspartate (NMDA) receptor mediates tau-induced toxicity leading to cell death⁶⁵. It is also well-established that glutamate-mediated neuronal hyperexcitability via NMDA receptors can cause seizures. Memantine, a non-competitive NMDA receptor antagonist, has been shown to treat both AD and seizures. It is widely used for the management of dementia because it improves cognition. At low doses, memantine has been shown to reduce hyper-excitability and seizures in animal models⁶⁶. These findings raise the possibility that memantine reduces cognitive decline and seizures by reducing hyper-excitabiltiy and tau-induced toxicity.

There is evidence to suggest that certain anti-seizure medications that stop seizure may also have a positive impact on the cognition of AD patients¹⁹. A case-control study of levetiracetam, lamotrigine, and phenobarbital utilization in AD patients with epilepsy over a 12-month period showed 72% seizure remission at one year, improved cognitive outcomes, and better MMSE performance, which was comparable to AD patients without epilepsy¹⁹.

In transgenic mice models of AD, suppression of epileptiform activity with anti-seizure medications (ASMs) improved network hyper-excitability in the hippocampus and cognitive deficits^67,68. An 8-week trial of lamotrigine among 11 AD patients WITHOUT history of epilepsy showed improved naming and recognition tasks at the end of the study period¹⁹. In a small-scale randomized controlled trial, treatment with low dose levetiracetam (250 mg/day) for 4 weeks improved executive function and spatial memory in AD patients who had subclinical epileptiform activity on extended EEG^19,69. However, these findings need to be confirmed in larger prospective and rigorous studies of longer duration.

We hypothesize that this impact of anti-seizure medications on improved cognition is likely because of the reduction of neuronal hyper-excitability, which in turn reduces epileptiform activity and seizures and leads to reduced accumulation and further spread of tau pathology.

Conclusion:

Evidence suggests several shared pathological features and a strong association between AD and MTLE. AD patients with MTLE or subclinical epileptiform activity appear to have accelerated and worse disease course with more significant neuronal loss and rapid cognitive and clinical decline^16,19,60 and shorter lifespans than AD patients without MTLE or subclinical epileptiform activity^4,11.

There is also evidence that these two disease processes may be characterized by excessive tau. Phosphorylated tau is frequently observed in epileptogenic tissue of post-surgical MTLE patients and in postmortem examination of deceased PwE^31,32. In addition, phosphorylated tau burden correlates with post-operative cognitive deficits among MTLE patients^31,32. Similarly among AD patients, it is the abnormally phosphorylated tau and NFT burden, rather than the Aβ load, especially in the entorhinal cortex and CA1 of the hippocampus, that is the predominant predictor of AD-related cognitive decline, cortical atrophy, and domain-specific cognitive changes^56-59. Phosphorylated tau is also associated with network dysfunction and neuronal hyperexcitability in a dose-dependent manner²².

Understanding the interplay between tau pathology, neuronal hyperexcitability, high co-occurrence of MTLE among AD patients, and factors that contribute to the development of epileptiform activity among AD patients is a crucial step toward understanding the overlap between the two diseases.

The findings above raise the possibility that AD patients with subclinical epileptiform activity or MTLE may represent a distinct disease process with the interplay between tau, neuronal hyperexcitablity and epileptiform activity at its core. It is imperative to understand this distinct disease process further to improve its clinical characterization, timely identification, and treatment. Another important unanswered question that needs to be explored in future studies is whether the more severe symptoms of AD with MTLE are due to the summation of two diseases or if it is a more distinct feature of AD beyond epilepsy.

Improved clinical characterization and a better understanding of how these disease overlap will not only help us reduce epileptogenicity among AD patients but also help us halt accelerated cognitive decline among these conditions and could potentially improve outcomes. Once the factors that contribute to epileptogenesis among AD patients have been identified, appropriate management strategies can be developed to reduce AD, more specifically, tau pathology as well as epileptiform activity among AD patients.

Future Directions:

Future studies are needed to establish a direct link between these two phenomena. Research is also needed to identify risk factors for developing MTLE among AD patients and identify predictors of cognitive decline and AD among MTLE patients. In addition, the link between lateral temporal lobe epilepsies and dementia remains unexplored and needs to be studied more thoroughly in future studies.

There is also a need to identify valid, reliable, and scalable biomarkers to predict better which AD patients are at the highest risk of developing epilepsy so that we can modify this risk with more personalized treatment approaches. Finally, there is a need to develop treatment strategies that will help reduce AD pathology and epileptogenesis among AD patients and PwE in order to halt the progression of cognitive decline and improve mortality outcomes.