SANAD II
The SANAD II study of the effectiveness and cost-effectiveness of levetiracetam, zonisamide, or lamotrigine for newly diagnosed focal epilepsy: an open-label, non-inferiority, multicentre, phase 4, randomised controlled trial
Year of Publication: 2021
Authors: Anthony Marson, Girvan Burnside, Richard Appleton, ..., on behalf of the SANAD II collaborators
Journal: The Lancet
Citation: Lancet 2021; 397: 1363–74
Link: https://doi.org/10.1016/S0140-6736(21)00247-6
PDF: https://www.thelancet.com/action/showPdf...%2821%2900247-6
PDF: Clinical Question
Are levetiracetam and zonisamide non-inferior to lamotrigine for long-term clinical effectiveness and cost-effectiveness as first-line treatment for patients with newly diagnosed focal epilepsy?
Study Overview
Objective
Levetiracetam - To assess the effectiveness and cost-effectiveness of levetiracetam and zonisamide versus lamotrigine in focal epilepsy, and levetiracetam versus valproate in generalized or unclassified epilepsy.
Study Summary
Lamotrigine and valproate showed superior outcomes compared to levetiracetam and zonisamide in their respective epilepsy types.
Intervention
Open-label, randomized controlled trial with stratified assignment: three-arm comparison in focal epilepsy and two-arm comparison in generalized/unclassified epilepsy.
Patients per Arm
Lamotrigine: 330; Levetiracetam: 332; Zonisamide: 328; Valproate: 260 (generalized only)
Bottom Line
Levetiracetam did not meet the definition of non-inferiority for time to 12-month remission compared with lamotrigine, and it was inferior for time to treatment failure. Zonisamide met non-inferiority criteria for 12-month remission but was also inferior for time to treatment failure. Neither levetiracetam nor zonisamide were cost-effective alternatives. These findings do not support the use of levetiracetam or zonisamide as first-line treatments for patients with focal epilepsy. Lamotrigine should remain a first-line treatment.
Major Points
- Levetiracetam did NOT meet non-inferiority criteria for 12-month remission (HR 1.18; 97.5% CI 0.95-1.47 includes the 1.329 margin)
- Zonisamide DID meet non-inferiority criteria for 12-month remission (HR 1.03; 97.5% CI 0.83-1.28)
- Per-protocol analysis showed lamotrigine SUPERIOR to both levetiracetam (HR 1.32) and zonisamide (HR 1.37)
- Lamotrigine was significantly less likely to fail than levetiracetam (HR 0.60; 95% CI 0.46-0.77) or zonisamide (HR 0.46; 0.36-0.60)
- Treatment failure was primarily due to adverse reactions, not inadequate seizure control
- Psychiatric adverse reactions were more common with levetiracetam (30%) and zonisamide (23%) than lamotrigine (13%)
- No significant difference in time to first seizure between groups (similar short-term efficacy)
- Lamotrigine dominated both comparators in cost-effectiveness analysis (higher net health benefit)
- Quality of life analyses showed levetiracetam and zonisamide associated with worse patient-reported outcomes
- 37 deaths occurred during the trial; no indication of higher death rate with any particular drug
Design
Study Type: Phase 4, multicentre, non-inferiority, open-label, randomised controlled trial
Randomization: 1
Blinding: Open-label. Participants and investigators were not masked and were aware of treatment allocation. Randomization was done using a secure, centrally controlled, 24-h web-based facility with minimisation program with random element utilising factors (centre, sex, number of previous seizures).
Enrollment Period: May 2, 2013 to June 20, 2017
Follow-up Duration: Minimum 2 years, maximum 7.5 years. Last participant visit October 17, 2019. Median follow-up: 462.5 days (lamotrigine), 449.5 days (levetiracetam), 447 days (zonisamide).
Centers: 65
Countries: United Kingdom
Sample Size: 990
Analysis: Intention-to-treat (ITT) primary analysis. Per-protocol (PP) analysis excluding major protocol deviations and participants subsequently diagnosed as not having epilepsy. Non-inferiority margin HR 1.329 (equates to 10% absolute difference). Cox proportional hazards regression. Fine and Gray competing risks model for treatment failure. One-sided alpha 0.0125 for primary comparisons; 97.5% CIs for primary outcome, 95% CIs for secondary outcomes. Power: 80% with one-sided alpha 0.025, 330 patients per group. SAS version 9.4.
Inclusion Criteria
- Age 5 years or older (no upper age limit)
- History of at least 2 unprovoked epileptic seizures requiring anti-seizure medication
- Clinical diagnosis of focal epilepsy with or without electroencephalogram
- Never treated with an anti-seizure medication except for emergency treatment in the previous 2 weeks
Exclusion Criteria
- Patients with provoked or acute symptomatic seizures only
- Patients currently taking anti-seizure medication
- Known progressive neurological disease (e.g., brain tumour)
Baseline Characteristics
Lamotrigine (n=330):
- Age, mean (SD; range), years: 40.1 (21.7; 5.1-91.9)
- Sex - Male, n (%): 186 (56%)
- Sex - Female, n (%): 144 (44%)
- Learning disability, n (%): 15 (5%)
- Febrile convulsions, n (%): 10 (3%)
- Acute symptomatic seizures, n (%): 6 (2%)
- History of epilepsy in primary relatives, n (%): 32 (10%)
- Neurological deficit, n (%): 12 (4%)
- Any previous or current neurological disorder, n (%): 57 (17%)
- Stroke or cerebrovascular, n (%): 17 (5%)
- Head injury, n (%): 4 (1%)
- Meningitis or encephalitis, n (%): 6 (2%)
- Temporal lobe epilepsy, n (%): 134 (41%)
- Frontal lobe epilepsy, n (%): 21 (6%)
- Parietal lobe epilepsy, n (%): 7 (2%)
- Occipital lobe epilepsy, n (%): 7 (2%)
- Focal epilepsy localisation not specified, n (%): 152 (46%)
- Total number of seizures, median (IQR): 6 (3-29)
- Days since first seizure, median (IQR): 333 (110-1090)
- Days since most recent seizure, median (IQR): 13 (3-41)
Levetiracetam (n=332):
- Age, mean (SD; range), years: 37.8 (20.1; 5.0-87.6)
- Sex - Male, n (%): 190 (57%)
- Sex - Female, n (%): 142 (43%)
- Learning disability, n (%): 16 (5%)
- Febrile convulsions, n (%): 19 (6%)
- Acute symptomatic seizures, n (%): 9 (3%)
- History of epilepsy in primary relatives, n (%): 35 (11%)
- Neurological deficit, n (%): 20 (6%)
- Any previous or current neurological disorder, n (%): 55 (17%)
- Stroke or cerebrovascular, n (%): 16 (5%)
- Head injury, n (%): 7 (2%)
- Meningitis or encephalitis, n (%): 5 (2%)
- Temporal lobe epilepsy, n (%): 110 (33%)
- Frontal lobe epilepsy, n (%): 21 (6%)
- Parietal lobe epilepsy, n (%): 8 (2%)
- Occipital lobe epilepsy, n (%): 12 (4%)
- Focal epilepsy localisation not specified, n (%): 165 (50%)
- Total number of seizures, median (IQR): 6 (3-22)
- Days since first seizure, median (IQR): 318 (119-985)
- Days since most recent seizure, median (IQR): 13 (3-35)
Zonisamide (n=328):
- Age, mean (SD; range), years: 39.9 (21.6; 5.0-89.1)
- Sex - Male, n (%): 185 (56%)
- Sex - Female, n (%): 143 (44%)
- Learning disability, n (%): 14 (4%)
- Febrile convulsions, n (%): 15 (5%)
- Acute symptomatic seizures, n (%): 4 (1%)
- History of epilepsy in primary relatives, n (%): 40 (12%)
- Neurological deficit, n (%): 12 (4%)
- Any previous or current neurological disorder, n (%): 51 (16%)
- Stroke or cerebrovascular, n (%): 14 (4%)
- Head injury, n (%): 7 (2%)
- Meningitis or encephalitis, n (%): 6 (2%)
- Temporal lobe epilepsy, n (%): 111 (34%)
- Frontal lobe epilepsy, n (%): 20 (6%)
- Parietal lobe epilepsy, n (%): 5 (2%)
- Occipital lobe epilepsy, n (%): 2 (1%)
- Focal epilepsy localisation not specified, n (%): 182 (55%)
- Total number of seizures, median (IQR): 6 (3-23)
- Days since first seizure, median (IQR): 328 (120-1097)
- Days since most recent seizure, median (IQR): 11 (3-34)
Total (N=990):
- Age, mean (SD; range), years: 39.3 (21.2; 5.0-91.9)
- Age <18 years, n (%): 177 (17.9%)
- Sex - Male, n (%): 561 (57%)
- Sex - Female, n (%): 429 (43%)
- Learning disability, n (%): 45 (5%)
- Febrile convulsions, n (%): 44 (4%)
- History of epilepsy in primary relatives, n (%): 107 (11%)
- Any previous or current neurological disorder, n (%): 163 (16%)
- Temporal lobe epilepsy, n (%): 355 (36%)
- Frontal lobe epilepsy, n (%): 62 (6%)
- Focal epilepsy localisation not specified, n (%): 499 (50%)
- Total number of seizures, median (IQR): 6 (3-24)
- Days since most recent seizure, median (IQR): 13 (3-36)
Arms
| Field | Control | Levetiracetam | Zonisamide |
|---|---|---|---|
| Intervention | Lamotrigine monotherapy. For participants aged ≥12 years: initial maintenance dose 50 mg morning and 100 mg evening (150 mg/day total). For children 5-12 years: 1.5 mg/kg twice daily. Subsequent dose adjustments according to routine clinical practice. | Levetiracetam monotherapy. For participants aged ≥12 years: initial maintenance dose 500 mg twice daily (1000 mg/day total). For children 5-12 years: 20 mg/kg twice daily. Subsequent dose adjustments according to routine clinical practice. | Zonisamide monotherapy. For participants aged ≥12 years: initial maintenance dose 100 mg twice daily (200 mg/day total). For children 5-12 years: 2.5 mg/kg twice daily. Subsequent dose adjustments according to routine clinical practice. |
| Duration | Long-term follow-up (minimum 2 years, maximum 7.5 years) | Long-term follow-up (minimum 2 years, maximum 7.5 years) | Long-term follow-up (minimum 2 years, maximum 7.5 years) |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Time to 12-month remission from seizures, calculated as days from randomisation to the first date at which a period of 12 months had elapsed without the patient having any seizures. Non-inferiority margin: HR 1.329 (equivalent to 10% absolute difference). | Primary | Lamotrigine: Median time to 12-month remission 516 days (95% CI 457-577) | Levetiracetam: 588 days (95% CI 472-706); Zonisamide: 530 days (95% CI 453-601) | Levetiracetam vs Lamotrigine: HR 1.18 (97.5% CI 0.95-1.47) - NON-INFERIORITY NOT MET (CI includes 1.329); Zonisamide vs Lamotrigine: HR 1.03 (97.5% CI 0.83-1.28) - NON-INFERIORITY MET | |
| Time to 12-month remission - Adjusted ITT analysis | Secondary | Lamotrigine reference | Levetiracetam: HR 1.13 (97.5% CI 0.91-1.41); Zonisamide: HR 1.01 (0.81-1.26) | ||
| Time to 12-month remission - Per-protocol analysis (Fine and Gray model) | Secondary | Lamotrigine reference | Levetiracetam: HR 1.32 (97.5% CI 1.05-1.66) - LAMOTRIGINE SUPERIOR; Zonisamide: HR 1.37 (1.08-1.73) - LAMOTRIGINE SUPERIOR | <0.025 | |
| Time to 24-month remission (ITT) | Secondary | Lamotrigine reference | Levetiracetam: HR 1.04 (95% CI 0.81-1.33); Zonisamide: HR 0.96 (0.75-1.23) | Not significant | |
| Time to first seizure (ITT) | Secondary | Lamotrigine reference | Levetiracetam: HR 1.07 (95% CI 0.89-1.29); Zonisamide: HR 1.04 (0.86-1.25) | Not significant | |
| Time to treatment failure for any reason (ITT) | Secondary | Lamotrigine reference | Levetiracetam: HR 0.60 (95% CI 0.46-0.77) - LAMOTRIGINE SIGNIFICANTLY BETTER; Zonisamide: HR 0.46 (0.36-0.60) - LAMOTRIGINE SIGNIFICANTLY BETTER | <0.001 | |
| Treatment failure at 2 years - absolute difference vs lamotrigine | Secondary | Lamotrigine reference | Levetiracetam: +16% more failures (95% CI 9-23); Zonisamide: +23% more failures (15-30) | ||
| Time to treatment failure due to adverse reactions (competing risks) | Secondary | Lamotrigine reference | Levetiracetam: HR 0.53 (95% CI 0.35-0.79) - SIGNIFICANT; Zonisamide: HR 0.37 (0.25-0.55) - SIGNIFICANT | <0.05 | |
| Time to treatment failure due to inadequate seizure control (competing risks) | Secondary | Lamotrigine reference | Levetiracetam: HR 0.67 (95% CI 0.45-1.01); Zonisamide: HR 0.76 (0.50-1.15) | Not significant | |
| Net health benefit at £20,000 per QALY (QALYs) - Base case | Secondary | Lamotrigine: 1.403 (97.5% CR 1.319-1.458) | Levetiracetam: 1.222 (1.110-1.283); Zonisamide: 1.232 (1.112-1.307) - LAMOTRIGINE DOMINATED | Probability of lamotrigine being cost-effective: 0.999 | |
| Total costs over 2 years (£) - Base case adjusted | Secondary | Lamotrigine: £4042 (97.5% CR 3626-4983) | Levetiracetam: £5104 (4450-6141); Zonisamide: £5400 (4659-6770) | ||
| QALYs - Base case adjusted | Secondary | Lamotrigine: 1.605 (97.5% CR 1.547-1.651) | Levetiracetam: 1.474 (1.393-1.523); Zonisamide: 1.502 (1.418-1.566) | ||
| Any adverse reaction - patients, n (%) | Adverse | Lamotrigine: 108/328 (33%) | Levetiracetam: 144/330 (44%); Zonisamide: 146/324 (45%) | ||
| Total adverse reaction events | Adverse | Lamotrigine: 251 events | Levetiracetam: 328 events; Zonisamide: 351 events | ||
| Psychiatric disorders - patients, n (%) | Adverse | Lamotrigine: 43/328 (13%) | Levetiracetam: 98/330 (30%); Zonisamide: 73/324 (23%) | ||
| Nervous system disorders - patients, n (%) | Adverse | Lamotrigine: 53/328 (16%) | Levetiracetam: 55/330 (17%); Zonisamide: 60/324 (19%) | ||
| General disorders (fatigue) - patients, n (%) | Adverse | Lamotrigine: 17/328 (5%) | Levetiracetam: 32/330 (10%); Zonisamide: 39/324 (12%) | ||
| Gastrointestinal disorders - patients, n (%) | Adverse | Lamotrigine: 25/328 (8%) | Levetiracetam: 22/330 (7%); Zonisamide: 26/324 (8%) | ||
| Skin and subcutaneous tissue disorders - patients, n (%) | Adverse | Lamotrigine: 24/328 (7%) | Levetiracetam: 12/330 (4%); Zonisamide: 21/324 (7%) | ||
| Rash events | Adverse | Lamotrigine: 22 events | Levetiracetam: 8 events; Zonisamide: 12 events | ||
| Serious adverse reactions | Adverse | Lamotrigine: 7 events in 2 patients | Levetiracetam: 1 patient; Zonisamide: 4 patients | ||
| Deaths | Adverse | Lamotrigine: 15 (4 possibly seizure-related) | Levetiracetam: 12 (2 possibly seizure-related); Zonisamide: 10 (2 possibly seizure-related). Total: 37 deaths | ||
| Pregnancies | Adverse | Lamotrigine: 11 pregnancies in 11 women (10 normal, 1 minor malformations) | Levetiracetam: 6 pregnancies in 5 women (5 normal, 1 termination); Zonisamide: 17 pregnancies in 14 women (8 normal, 8 miscarriages in 5 women, 1 termination) |
Subgroup Analysis
Children aged <16 years (n=155): Levetiracetam had the highest net health benefit (1.307 QALYs vs 1.297 for lamotrigine and 1.277 for zonisamide), but limited by small sample size and principally due to a single participant with atypical medical journey. Adults and adolescents aged ≥16 years (n=835): Lamotrigine dominated with net health benefit of 1.420 QALYs vs 1.207 (levetiracetam) and 1.227 (zonisamide). Only 177 (17.9%) participants were aged <18 years, limiting applicability to children. QOL analyses: For adults, lamotrigine associated with better profile on self-reported measures; levetiracetam showed negative treatment effects for anxiety, depression, stigma, epilepsy impact, and overall QOL compared to lamotrigine.
Criticisms
- Open-label design - neither participants nor investigators were masked to treatment allocation
- Seizure data collected using diaries and clinic reports may have missed unreported seizures, potentially affecting treatment decisions
- Only 177 (17.9%) participants were younger than 18 years, limiting applicability to children
- Paediatrician lack of experience with zonisamide (not licensed as monotherapy in children) may have affected recruitment
- No randomised trial data to inform choice of initial maintenance doses for any of the drugs
- Low return rate for quality of life questionnaires (49.8% at baseline and at least one other timepoint)
- Participants who returned questionnaires were a mean of 10 years older than those who did not
- Economic analysis limited by poor questionnaire return and inclusion of free-text questions
- EQ-5D-3L-Y utility valuation for children required application of adult tariff
- Did not assess other newer anti-epileptic drugs such as lacosamide (now licensed as monotherapy) or perampanel
- UK-only study may limit international generalizability
- Small number of pregnancies limits conclusions about pregnancy outcomes with zonisamide (8 miscarriages)
Funding
National Institute for Health Research (NIHR) Health Technology Assessment Programme (project reference 09/144/09). Trial co-sponsored by the University of Liverpool and the Walton Centre NHS Foundation Trust. The funder had no role in study design, data collection, analysis, interpretation, or writing of the report.
Based on: SANAD II (The Lancet, 2021)
Authors: Anthony Marson, Girvan Burnside, Richard Appleton, ..., on behalf of the SANAD II collaborators
Citation: Lancet 2021; 397: 1363–74
Content summarized and formatted by NeuroTrials.ai.