Seizure Prophylaxis After a Coiled Aneurysm: Worth Starting at All
A gardener's ruptured aneurysm is secured and she hasn't had a single seizure. The question is whether starting a drug to prevent one she may never have is actually the safer choice.
M.H., a 46-year-old woman, was weeding a flowerbed at the garden center where she works part-time when the headache hit — sudden, the worst of her life by her own description, followed by seconds of witnessed loss of consciousness that resolved before EMS arrived. Imaging found diffuse subarachnoid blood from a ruptured anterior communicating artery aneurysm, secured the same day via endovascular coiling rather than open clipping. A full day later, she has had no witnessed convulsion, no epileptiform activity on continuous EEG, and no parenchymal hematoma or cortical involvement on her scans — findings that matter directly to today's question, because they describe a genuinely lower-risk seizure profile than a comparable patient with cortical blood or a surgical craniotomy incision would carry.
The instinct to start a prophylactic antiepileptic anyway is understandable, but the evidence behind that instinct, specifically for phenytoin, runs the other direction: Rosengart and colleagues' 2007 retrospective study of SAH patients found phenytoin exposure independently associated with worse cognitive outcome at follow-up, a real cost attached to a drug whose seizure-prevention benefit in this population was never established by a randomized trial in the first place. The current AHA/ASA subarachnoid hemorrhage guideline reflects that same evidence gap, recommending against routine prophylactic antiepileptic use beyond the immediate perioperative period rather than endorsing it as standard care. She has no cortical blood, no seizure, and a coiled rather than clipped aneurysm — a seizure-risk profile that argues for treating the guideline's default of no routine prophylaxis as the actual answer here, not a starting point to be talked out of with a reflexive short course.
She has also asked, more than once since regaining her bearings, whether the headache means she'll need to stay off her feet at the garden center for good — a question the team has deferred to her outpatient follow-up, since her vasospasm-monitoring window and any activity restrictions are a separate conversation from today's seizure-prophylaxis decision. Her only other medical history is a remote tubal ligation, otherwise unremarkable, which leaves this decision resting on the aneurysm and its treatment alone rather than any competing comorbidity.
A prophylactic drug for a risk nobody has quantified for her
I know the guideline leans against routine prophylaxis, and I'm not arguing for a full course. But she's one day out from an aneurysm rupture and a coiling procedure, and I'd feel better with some coverage through the immediate perioperative window — a short course, not an open-ended one.
That's a reasonable middle position, and I don't think it's wrong to want some margin this early. My concern is specifically about which drug fills that role. If we're going to give anything, it can't be phenytoin — Rosengart's data found real cognitive cost associated with it in exactly this population, without a randomized trial ever establishing the seizure-prevention benefit that would justify taking that cost on.
Then the actual choice in front of us is levetiracetam for a defined short window, not phenytoin at all — phenytoin was never seriously on the table once its own outcome data is named directly. A 72-hour course, reassessed and stopped if she remains seizure-free with a clean EEG, gives the neurosurgeon's perioperative caution a real answer without carrying the specific drug whose cost is already documented in this exact population.
Levetiracetam ran for 72 hours with no seizure activity on continuous EEG, and it was discontinued on schedule without incident. She remained seizure-free through the remainder of her admission.
The neurosurgeon's original instinct for some perioperative coverage and the neurointensivist's guideline-grounded caution both got a real answer rather than either simply overriding the other — the short, time-limited course satisfied the first without reintroducing the drug-specific risk the second was concerned about. No further AED was recommended at discharge.