News|Articles|November 18, 2026

Levetiracetam Misses Primary Endpoint in Early Alzheimer Trial

Author(s)Marco Meglio
Listen
0:00 / 0:00

Key Takeaways

  • A randomized 3-period crossover trial (placebo, 125 mg BID, 500 mg BID) in early AD showed no significant omnibus effect on an NTB global composite (P=.33).
  • Low-dose levetiracetam produced the largest numerical signal in memory (difference 0.20; P=.10), whereas executive function showed no favorable trend (omnibus P=.56).
SHOW MORE

A phase 2a crossover trial testing low-dose and high-dose levetiracetam in early Alzheimer disease did not significantly improve global cognition versus placebo.

Researchers tested low-dose and high-dose levetiracetam (LEV) against placebo in adults with early AD in a double-blind crossover trial presented as a poster at the 2026 Clinical Trials on Alzheimer's Disease (CTAD) conference, held November 17-20 in Las Vegas, Nevada. All told, neither dose significantly improved global cognition over placebo, though the trial stratified participants by brain excitability to test who might respond.1

The trial, led by Stephanie S. Buss, MD, cognitive neurologist and Assistant Professor of Neurology at Harvard Medical School, and others, screened 123 adults and enrolled 43; 31 were randomized to 1 of 6 treatment sequences covering placebo, low-dose LEV (125 mg twice daily), and high-dose LEV (500 mg twice daily), each given for 4 weeks and separated by 4-week washouts.

Eligible participants had biomarker-confirmed AD with a Clinical Dementia Rating (CDR) of 0.5 to 1. Thirty participants completed at least 1 treatment period and were included in the modified intention-to-treat analysis; 23 completed all 3 periods. COVID-19 disruptions reduced enrollment below the original target, leaving the trial powered to detect a moderate effect.

Participants averaged 70.6 years old, and 83% had mild cognitive impairment (MCI) due to AD rather than mild dementia; baseline Mini-Mental State Examination (MMSE) scores averaged 26.1. Investigators stratified participants into tertiles of baseline cortical excitability using transcranial magnetic stimulation combined with electroencephalography (TMS-EEG), in which a figure-of-8 coil delivered single pulses to the inferior parietal lobe while a 64-channel EEG recorded the evoked response, building on source-localization methods the same research group recently validated in AD.1,2

On the primary endpoint, a global composite from a Neuropsychological Test Battery (NTB) spanning executive function and memory tests, LEV did not produce a statistically significant omnibus effect across the 3 arms (P =.33); the observed effect size (Cohen f=0.22) fell below what the trial was powered to detect. In prespecified pairwise contrasts against placebo, low-dose LEV numerically improved the NTB composite by 0.12 points (95% CI, -0.04 to 0.28; P=.15) and high-dose LEV by 0.03 points (95% CI, -0.12 to 0.18; P =.70).1

The pattern was most pronounced on the memory composite, where low-dose LEV improved scores by 0.20 points versus placebo (95% CI, -0.04 to 0.43; P =.10), a trend that did not reach significance, while high-dose LEV showed a smaller, nonsignificant difference of 0.07 points (95% CI, -0.15 to 0.30; P =.52). The executive function composite showed no comparable trend in either direction (omnibus P =.56).

Several prespecified subgroup analyses suggested LEV's effects varied by baseline characteristics, though most comparisons were underpowered given the small sample. Baseline TMS-EEG parietal excitability moderated response to treatment (tertile interaction P=.05): participants in the most excitable tertile trended toward benefit from high-dose LEV (difference vs placebo, 0.21 points; P =.085), while those in the least excitable tertile trended toward benefit from low-dose LEV (0.23 points; P =.103) and toward harm from high-dose LEV (-0.21 points; P =.16).

Clinical stage and baseline disease severity each also moderated the response (both P =.04), and participants without an APOE4 allele showed numerically larger effects than APOE4 carriers, though that subgroup included only 4 people.

READ MORE: Roche Launches Phase 2 TARBELA-AD Trial of NLRP3 Inhibitor in Alzheimer Disease

LEV was well tolerated across both doses, with no differences in neuropsychiatric symptoms measured by the Neuropsychiatric Inventory Questionnaire (NPI-Q) and no increase in suicidal ideation on the Columbia-Suicide Severity Rating Scale (C-SSRS), which showed its greatest severity during the placebo period. Peak blood LEV concentration was not linearly associated with cognitive change (slope, -0.005 per µg/mL; 95% CI, -0.015 to 0.004; P =.31), though the data hinted at a nonlinear, dose-dependent relationship consistent with the trial's prespecified hypothesis of an inverted-U response curve.

Levetiracetam's winding path toward an Alzheimer indication

The rationale for testing LEV in AD dates back more than a decade. Subclinical epileptiform activity, abnormal electrical discharges often undetectable without specialized monitoring, is more common in AD than previously recognized and has been linked to faster cognitive decline, giving researchers a mechanistic reason to test an antiseizure drug in this population.3

A 2012 proof-of-concept study first linked low-dose LEV to reduced hippocampal hyperactivity and improved performance on a memory task in adults with amnestic MCI, establishing the rationale that lower doses, rather than standard antiseizure doses, might best normalize AD-related network dysfunction without further impairing cognition.4

That rationale was tested more directly in a 34-patient phase 2a crossover trial, which found that 125 mg of LEV twice daily did not improve overall cognition across the full AD cohort, but in an exploratory analysis of the 13 participants with confirmed epileptiform activity, LEV improved performance on a test of executive function and on a virtual-reality route-learning task.5 The new trial's dosing and crossover structure closely mirror that earlier study, though it did not require confirmed epileptiform activity for enrollment, relying instead on TMS-EEG excitability measures to stratify participants.

The most recent large-scale test of this approach, the HOPE4MCI trial, evaluated AGB101, an extended-release formulation of low-dose LEV (220 mg once daily), in 164 adults with amyloid-confirmed MCI due to AD over 78 weeks. That trial also missed its primary endpoint, change on the CDR-Sum of Boxes, across the full cohort, though a prespecified subgroup analysis numerically favored APOE4 noncarriers, a pattern echoed in the new trial's own subgroup findings.6

Taken together, the authors characterized their results as inconclusive rather than negative, given the trial's reduced sample size. They reported that the overall pattern favored low-dose LEV, predominantly in the memory domain, that treatment effects differed by cortical excitability and disease severity, and that APOE4 noncarriers showed a larger effect size than carriers. The findings, they concluded, support designing an adequately powered trial of LEV in early symptomatic AD that stratifies participants by baseline cortical excitability.1

For more CTAD 2026 coverage, click here.

REFERENCES
1. Buss SS, Passera B, Ozdemir R, et al. Low-dose levetiracetam to treat cortical hyperexcitability in early Alzheimer's disease: a randomized, double-blind phase 2a crossover trial. Presented at: 2026 Clinical Trials on Alzheimer's Disease Conference; November 16-19, 2026; Boston, MA.
2. Ozdemir RA, Passera B, Fried PJ, et al. Neurophysiological signatures of default mode network dysfunction and cognitive decline in Alzheimer's disease. Sci Adv. 2025;11(48):eadt8991. doi:10.1126/sciadv.adt8991
3. Vossel KA, Ranasinghe KG, Beagle AJ, et al. Incidence and impact of subclinical epileptiform activity in Alzheimer's disease. Ann Neurol. 2016;80(6):858-870. doi:10.1002/ana.24794
4. Bakker A, Krauss GL, Albert MS, et al. Reduction of hippocampal hyperactivity improves cognition in amnestic mild cognitive impairment. Neuron. 2012;74(3):467-474. doi:10.1016/j.neuron.2012.03.023
5. Vossel K, Ranasinghe KG, Beagle AJ, et al. Effect of levetiracetam on cognition in patients with Alzheimer disease with and without epileptiform activity: a randomized clinical trial. JAMA Neurol. 2021;78(11):1345-1354. doi:10.1001/jamaneurol.2021.3310
6. Mohs R, Bakker A, Rosenzweig-Lipson S, et al. The HOPE4MCI study: a randomized double-blind assessment of AGB101 for the treatment of MCI due to AD. Alzheimers Dement (N Y). 2024;10(1):e12446. doi:10.1002/trc2.12446

Related to this article