Reviews/Epilepsy Management/Levetiracetam non-inferior to lamotrigine
Asm Efficacy · EP-ASM-01 · v2
Borderline: within 4 points of the Doubtful/Refuted boundary, so a re-grade could land either side.
Only one study in this corpus, the SANAD II trial [3], was actually built to test whether levetiracetam is 'non-inferior' to lamotrigine in focal epilepsy, meaning no meaningfully worse within a pre-agreed amount. In that trial levetiracetam failed to clear the bar, and when patients who stayed on their assigned drug were analysed, lamotrigine came out ahead on both seizure remission and the rate at which treatment had to be stopped [3]. Several real-world observational studies [1][6][13] found the two drugs roughly similar for seizure control, and one even favoured levetiracetam [13], while a persistence study found them nearly identical [8]. But these observational studies were not blinded, patients were not randomly assigned, and they do not test the formal non-inferiority bar. Because the one trial designed for this exact question did not support the claim and pointed toward lamotrigine, the panel landed against the claim as a formal non-inferiority statement.
The purpose-built non-inferiority RCT [3] found levetiracetam did not meet the non-inferiority criterion versus lamotrigine for time to 12-month remission (ITT hazard ratio 1.18, 97.5% CI 0.95 to 1.47; margin 1.329), with lamotrigine superior per-protocol (HR 1.32, 95% CI 1.05 to 1.66) and for treatment failure (HR 0.60, 95% CI 0.46 to 0.77). Observational cohorts point the other way on crude comparability: [6] found no significant one-year seizure-free difference and similar three-year rates (LTG 39.8% vs LEV 41.4%), and [13] found levetiracetam superior for seizure freedom (72.44% vs 66.49%, p=0.006). Persistence data [8] were near-identical at one year (LTG 57.5% vs LEV 58.3%), and in elderly new-onset focal epilepsy [1] levetiracetam retention (61.5%) was close to lamotrigine (55.6%), which 'did not differ significantly from either comparator'. The corpus thus shows rough real-world comparability but a negative result on the one formal margin test. The contested point: The division is between evidentiary objects: the randomised, margin-based test [3] says levetiracetam is not non-inferior and lamotrigine is better per-protocol, while the unblinded observational cohorts [1][6][13] and persistence data [8] show the two drugs roughly comparable. The disagreement is whether crude point-estimate similarity in confounded real-world data can stand in for a pre-specified non-inferiority margin, which it cannot.
All 16 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Epilepsy Management, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
Non-inferiority is a formal statistical claim against a pre-specified margin; only [3] tests it and LEV failed to meet it (upper 97.5% CI 1.47 > margin 1.329). Observational comparability answers a weaker, different question and cannot establish non-inferiority.
[3] per-protocol shows LTG significantly superior to LEV for 12-month remission (HR 1.32, 1.05-1.66) and treatment failure (HR 0.60, 0.46-0.77) — this is positive evidence against the claim, not mere unproven non-inferiority.
Where it remains uncertain. Observational cohorts [6][13] and persistence data [8] show the drugs roughly comparable, and [13] even favours levetiracetam (72.44% vs 66.49%, p=0.006), but none tests a pre-specified non-inferiority margin.
Aleatory
The ITT point estimate in [3] (HR 1.18) carries a 97.5% CI of 0.95 to 1.47 that spans both no difference and clinically meaningful inferiority, so random variation leaves the true difference imprecisely located.
Epistemic
A single purpose-built non-inferiority RCT [3] addresses the exact claim; no second randomised margin test exists in the corpus to confirm or overturn its result.
The load-bearing papers from a corpus of 16, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Countervailing RCT signal of LEV/LTG comparability on retention, but in elderly only and not a formal NI test.
“At week 58, the retention rate for LEV was significantly higher than for CR-CBZ (61.5% vs. 45.8%, p = 0.02), and similar to LTG (55.6%).”
Panel auditInternal validity: high · scope of inference: Elderly (aged ≥60) patients with newly diagnosed focal epilepsy initiating monotherapy.Key limitations: elderly-only population · LEV vs LTG not the primary comparison · retention not seizure-control primary endpoint
Observational cohort shows LEV and LTG comparable on 1-year seizure freedom in focal epilepsy.
“no significant difference was observed among the three drugs for the one-year seizure-free rate.”
Panel auditInternal validity: medium · scope of inference: Patients aged 1-70 with newly diagnosed focal epilepsy on monotherapy in a real-world setting.Key limitations: observational, non-randomized · LEV vs LTG not the primary contrast · wide age range
Indirect persistence signal of comparability, not efficacy non-inferiority.
“In ASM-naïve patients, adherence and 1-year persistence were almost identical in the lacosamide, lamotrigine, and levetiracetam groups.”
Panel auditInternal validity: medium · scope of inference: Adults with focal epilepsy initiating ASMs in a Japanese insurance claims database.Key limitations: persistence proxy not efficacy · no p-value for LEV vs LTG · claims-data confounding
Pivotal purpose-built non-inferiority RCT directly refutes the claim: LEV did not meet the 1.329 margin versus LTG.
“The SANAD II findings do not support the use of levetiracetam or zonisamide as first-line treatments in focal epilepsy.”
Panel auditInternal validity: high · scope of inference: Patients aged ≥5 years with focal epilepsy requiring anti-seizure medication, UK NHS outpatient setting.Key limitations: unblinded design · pragmatic setting may affect dosing decisions
4 of 16 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
The pivotal, directly-on-point non-inferiority RCT refutes the claim and favours lamotrigine, outweighing lower-grade observational comparability.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
A blinded non-inferiority RCT in general focal epilepsy. The only margin test [3] was pragmatic and unblinded, which its authors note could bias dosing, treatment-failure and adverse-reaction decisions. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A double-blind randomised non-inferiority trial of levetiracetam versus lamotrigine with a pre-specified margin and time-to-12-month-remission endpoint.
Open question: Would levetiracetam meet a non-inferiority margin versus lamotrigine under blinded randomised conditions in general focal epilepsy? Why it is still open: The only margin test in the corpus [3] was unblinded and its authors flag that this could have biased the result.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19