This is a long-brewing topic on which I have changed my mind (a few times!), based on the evidence as it rolls out.
- Does Alzheimer’s disease cause patients to have seizures?
- If it does, how often should we treat with anti-seizure medication to improve the cognition or behavior of someone with Alzheimer’s disease?
Because the pathology of Alzheimer’s disease affects the cortex of the brain (AKA gray matter), there is always a potential for abnormal cortical electrical activity to arise and propagate over areas of the brain, that might be focal (e.g., causing momentary speech arrest) or generalized throughout the brain. We have an easier time of recognizing generalized seizures (loss of consciousness, often with major involuntary jerking or shaking as you see in the movies). Those patients, regardless of whether they have Alzheimer’s disease (AD), do get treated so that we can prevent additional generalized seizures, which can cause injury but are not typically fatal. The literature contains reports of anywhere from 20-50% of AD patients at given centers showing seizure activity. But anti-seizure medications (AKA anti-epileptic drugs) can have adverse effects on cognition, and the risk:benefit ratio has been difficult to distinguish in this seizure context. For a long time, I did not treat with anti-seizure medication unless a generalized seizure had occurred or a patient’s severe agitation warranted treatment with valproic acid. Over time, some patients slept better, avoided seizures and seemed a bit more clear to their caregivers; others progressed into end stages of AD and it was hard to tell if there were further seizures unless convulsions were observed by caregivers. [Not all seizures manifest with convulsions; it depends on the part of the brain with abnormal electrical activity and synchronization.] Here’s a 2016 review paper.
Then clinical trials did not show clear benefit of valproic acid for agitation, so I shied away from that second indication for prescribing anti-seizure medication.
From 2014-2015, I was reading EEGs in my small, Hawai’ian private practice, and I saw with my own skeptical eyes that many of my patients with Alzheimer’s disease did have indications of seizure activity, even if none of them had a seizure during the 1.5 hour recording session (e.g., the tracing shown above).
It was hard to tell from the few patients who could tolerate regular administration of an anti-seizure medication (and believe me, we tried several in sequence on the off-chance that we could clear cognition, even if a little bit), if this improved their lot.
The latest study I’ve seen, which triggered writing this post, seems to back up what I’ve experienced less formally in clinic: Keith Vossel at UCSF et al. published work showing that patients with early stages of AD are 4 times more likely than age-matched healthy folks to show that evidence of seizure activity on EEG over a 3.3-year period. And those 42.4% of patients in the study with the EEG abnormalities proceeded into more severe stages of AD more quickly over the course of the study than the 58% of AD patients in the study with normal EEGs. My takehome messages from this important paper are:
- there may be a role for doing at least one EEG on each patient with AD, even if in an early stage of illness
- it might not be worth doing that from a cost-benefit point of view unless we can determine a tolerable anti-seizure medication for this population of patients first. As an example, despite warnings that levetiracetam can cause irritability in older patients, there has been some work to show efficacy as an anti-agitation agent.