On a DOAC, lamotrigine or lacosamide were linked to fewer clots and deaths than other antiseizure drugs

Design

Retrospective target-trial emulation with propensity matching

Population

9,529 adults with epilepsy on a DOAC starting antiseizure monotherapy

Primary outcome

Thromboembolic events, major bleeding and all-cause mortality

Effect

Levetiracetam vs lamotrigine/lacosamide: thromboembolism HR 1.98 (95% CI 1.37–2.87); mortality HR 1.60

Many antiseizure drugs interact with direct oral anticoagulants (DOACs) — enzyme inducers lower DOAC levels, and other drugs may interact through transporters or affect bleeding. This study used electronic health records from 165 healthcare organisations to emulate a trial in 9,529 adults with epilepsy on a DOAC who started antiseizure monotherapy, comparing each drug group with lamotrigine or lacosamide after propensity matching.

Against lamotrigine or lacosamide, levetiracetam was associated with more thromboembolic events (HR 1.98, 95% CI 1.37–2.87) and higher mortality (HR 1.60, 95% CI 1.23–2.08), with similar bleeding. Strong enzyme inducers such as carbamazepine and phenytoin were associated with more thromboembolism (HR 1.55, 95% CI 1.02–2.36) and less major bleeding — consistent with lower DOAC levels. Valproate was associated with higher mortality (HR 1.49) and more intracranial bleeding (HR 3.01, 95% CI 1.45–6.26). In patients on warfarin, thromboembolic differences were close to null.

The levetiracetam finding is unexpected, since it is usually considered interaction-free, and may reflect confounding by who receives it — often older, frailer patients after stroke. Even so, the direction of the evidence supports lamotrigine or lacosamide as first choices when a new antiseizure drug is started in someone on a DOAC, and avoiding strong enzyme inducers. It is observational and does not justify switching a patient who is stable.

• When starting an antiseizure drug in a patient on a DOAC, consider lamotrigine or lacosamide first.
• Avoid strong enzyme inducers such as carbamazepine and phenytoin with DOACs where possible.
• Use caution with valproate in anticoagulated patients, given the intracranial bleeding signal.
• Do not switch a stable, seizure-free patient on this evidence alone; discuss the trade-offs.
• Review anticoagulant choice and dosing whenever an antiseizure drug is started or stopped.

Why it matters

Antiseizure drug choice in anticoagulated patients may be a modifiable contributor to stroke and death, not just to seizure control.

Don't overread it

The levetiracetam signal comes from observational data and may reflect who receives it rather than the drug itself.

The statistics, in plain English

Propensity matching makes groups similar on recorded factors, but cannot balance what was not recorded, such as frailty or stroke severity. Hazard ratios near 2 are large enough to take seriously, but not proof of cause.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.