Levetiracetam, a widely used antiseizure medication (ASM) often favored for its low potential for drug interactions, may warrant greater caution when prescribed with direct oral anticoagulants (DOACs), after new observational data linked the combination to a nearly twofold higher risk for thromboembolic events.

In a retrospective study of more than 9000 adults with epilepsy receiving DOACs, levetiracetam was associated with nearly twice the thromboembolic risk compared with lamotrigine or lacosamide. Strong enzyme-inducing ASMs, including carbamazepine and phenytoin, were associated with a 55% higher thromboembolic risk compared with the reference drugs.

Levetiracetam and valproate were also associated with higher mortality, whereas valproate was tied to a greater risk for intracranial major bleeding.

“The magnitude of the levetiracetam signal surprised us,” Kai Michael Schubert, MD, PhD, co-investigator, Department of Neurology, Clinical Neuroscience Center, University Hospital and University of Zurich in Zurich, Switzerland, told Medscape Medical News. “The association was strong and consistent enough that I would no longer regard levetiracetam as an automatically low-interaction choice during DOAC therapy; this combination deserves caution and active review.”

The study was published online on August 17 in JAMA Neurology. 

A Complex Drug Combination

DOACs are used for stroke prevention in atrial fibrillation and to treat venous thromboembolism, including in patients with epilepsy. Yet concurrent ASMs can impair their efficacy because some agents alter drug metabolism or transport.

Carbamazepine, phenytoin, phenobarbital, and other strong enzyme-inducing ASMs can increase CYP3A4 and P-glycoprotein activity, which, in turn, can lower DOAC exposure and anticoagulant efficacy.

Although levetiracetam is considered a low-interaction, nonenzyme-inducing alternative, prior observational studies have left it unclear whether the thromboembolic signal is drug specific.

To compare thromboembolic, bleeding, and mortality risks in adults with epilepsy receiving DOACs, investigators used de-identified electronic health record data from 165 healthcare organizations in the TriNetX Global Collaborative Network. They compared moderate and strong enzyme-inducing ASMs, valproate, and levetiracetam, with lamotrigine or lacosamide.

The retrospective cohort study included 9529 adults with epilepsy (mean age, 63 years; 51% women), with 5473 adults receiving levetiracetam, 1395 receiving strong enzyme-inducing ASMs, 1006 receiving valproate, 382 receiving moderate enzyme-inducing ASMs, and 1273 receiving lamotrigine or lacosamide.

The investigators used 1:1 propensity-score matching and a 90-day landmark to reduce baseline differences and potential reverse causation.

Levetiracetam Raises New Concerns

Levetiracetam was linked to a nearly twofold increased risk for thromboembolic events (hazard ratio [HR], 1.98) and associated with higher all-cause mortality (HR, 1.60), with no significant difference in major bleeding risk (HR, 1.01), compared with lamotrigine or lacosamide.

Strong enzyme-inducing ASMs carried a higher thromboembolic risk (HR, 1.55) but a lower risk for major bleeding (HR, 0.62). Valproate was tied to a higher risk for mortality (HR, 1.49) and a threefold increased risk for intracranial major bleeding (HR, 3.01). Moderate enzyme-inducing ASMs did not show any clear differences with thromboembolism (HR, 1.15), major bleeding (HR, 0.69), or mortality (HR, 1.17).

The differences in thromboembolic risk were modest but potentially meaningful. Thromboembolic events occurred in 5.5% of patients receiving levetiracetam, 5.8% of those receiving strong enzyme-inducing ASMs, 4.7% of those receiving valproate, and 4.6% of those receiving moderate enzyme-inducing ASMs, compared with 3.7% of patients receiving lamotrigine or lacosamide.

The investigators estimated that 124 of 435 thromboembolic events, or about 28%, potentially could have been prevented if all patients had received either lamotrigine or lacosamide. However, this modeled estimate does not mean that switching ASM therapy would prevent 28% of events in routine practice, they cautioned.

The risk for thromboembolic events was similar across all ASM groups for patients on vitamin K antagonists, unlike higher risks seen in DOAC users. This pattern is consistent with, although does not prove, a DOAC-specific effect, the investigators noted.

“ASM choice should be considered a modifiable part of anticoagulation safety,” Schubert said.

The findings do not prove that levetiracetam directly causes thromboembolism, or that switching agents will lower risk.

Since levetiracetam is not a classic cytochrome P450 inducer and clinically relevant induction of human P-glycoprotein has not been established, Schubert noted, “the mechanism remains unknown and requires dedicated research.”

‘Hypothesis-Strengthening Not Practice-Changing’

The findings are important, but they should not prompt clinicians to rush to change treatment plans, cautioned Sean Hennessy, PharmD, PhD, professor of epidemiology and of systems pharmacology and translational therapeutics at the Perelman School of Medicine at the University of Pennsylvania in Philadelphia.

“I would view these findings as hypothesis-strengthening rather than necessarily practice-changing,” Hennessy, who was not involved in the study, told Medscape Medical News.

He pointed to a potential disconnect in the findings: If levetiracetam were reducing DOAC exposure through a pharmacokinetic interaction, he would expect bleeding risk to fall as well. Alternatively, a pharmacodynamic effect should produce a similar signal among patients taking vitamin K antagonists, which was not observed.

The increase in all-cause mortality also raises the possibility of residual confounding, he said, because “thromboembolic events account for only a small fraction of deaths.” However, the largely null findings for outcomes such as falls and pneumonia argue against a simple explanation based on “channeling of frailer patients toward levetiracetam.”

Michael Gelfand, MD, PhD, associate professor of clinical neurology at the University of Pennsylvania, who was also not involved in the study, similarly raised the possibility of residual confounding.

“Levetiracetam is often the drug of choice for acutely ill ER [emergency room] or ICU patients,” Gelfand told Medscape Medical News. “The increased mortality identified in the new study wouldn’t be explained by thromboembolism, but it would be explained by a sicker population.”

The new findings differ from those of 2024 JAMA Neurology study by Gelfand, Hennessy, and colleagues, which found no significant differences in thromboembolic risk with enzyme-inducing ASMs but a lower risk for major bleeding.

Differences in propensity-score matching could help explain the discrepancy because patients who received enzyme-inducing ASMs may have remained sicker after matching in the current study, Gelfand explained.

He also noted that the prior study included a prespecified sensitivity analysis excluding levetiracetam from its nonenzyme-inducing reference group, which did not change its findings.

Although strong enzyme-inducing ASMs remain best avoided with DOACs, Gelfand said, levetiracetam remains a reasonable alternative, albeit one that may warrant closer scrutiny.

“The new findings suggest caution, but I would nevertheless continue to consider levetiracetam a good alternative for patients on DOACs,” he said. “Patients doing well on levetiracetam should not rush to switch based on these findings.”

Disclosure information for study authors is available in the original study publication. Hennessy and Gelfand reported no relevant financial disclosures.