ASM Selection & Treatment
ASM Selection & Treatment Principles
What Do You Need to Know?
- Focal epilepsy first-line: lamotrigine (SANAD I + II); alternatives: LEV, CBZ, OXC, lacosamide, cenobamate
- Generalized epilepsy first-line: valproate (SANAD I + II); use LTG or LEV in women of childbearing potential
- Broad-spectrum ASMs (VPA, LEV, LTG, TPM, ZNS, clobazam, perampanel) are safe for all seizure types; narrow-spectrum Na+ channel blockers may worsen IGE
- Seizure freedom probability: ~50% with first ASM, ~13% with second, 2–5% each subsequent; cumulative ~64%
- Drug-resistant epilepsy (ILAE 2010): failure of 2 tolerated, appropriately chosen and used ASMs → refer for surgical evaluation
- Synergistic combination with strongest evidence: LTG + VPA (but VPA doubles LTG levels — dose adjust)
- Withdrawal: consider after 2–5 years seizure-free; recurrence ~30–40% overall; JME >80–90% relapse → generally lifelong treatment
Broad-Spectrum vs. Narrow-Spectrum ASMs
Classification by Spectrum
| Spectrum | ASMs | Indications | Key Caveat |
|---|---|---|---|
| Broad-spectrum | VPA, LEV, LTG, TPM, ZNS, clobazam, perampanel | Safe for ALL seizure types; preferred when classification uncertain | LTG may worsen myoclonus in some IGE; TPM ineffective for absence |
| Narrow-spectrum (focal) | CBZ, OXC, ESL, PHT, lacosamide, GBP, PGB | Focal epilepsy only | May worsen absence, myoclonic, atonic — AVOID in IGE |
| Narrow-spectrum (absence) | Ethosuximide | Absence seizures only | No efficacy for other seizure types; first-line for CAE without GTC |
- Rule: if seizure type uncertain → always choose broad-spectrum ASM
- Na+ channel blockers exacerbate generalized absence/myoclonic seizures via enhanced thalamocortical synchronization
- Gabapentinoids (GBP, PGB) may worsen myoclonus and absence seizures
ASM Selection by Seizure Type & Syndrome
Master Selection Table
| Seizure Type / Syndrome | First-Line | Alternatives | Key Evidence / Notes |
|---|---|---|---|
| Focal seizures | LTG | LEV, CBZ, OXC, lacosamide, cenobamate | SANAD I + II: LTG best for time to remission and treatment failure |
| Generalized tonic-clonic | VPA | LTG, LEV | SANAD I + II: VPA superior; LTG/LEV preferred in women of childbearing potential |
| Absence (CAE without GTC) | Ethosuximide | VPA, LTG | SANAD: ETX = VPA > LTG for absence freedom; ETX preferred (fewer AEs) |
| Absence (with GTC) | VPA | LTG, LEV | ETX ineffective for GTC → VPA covers both seizure types |
| Myoclonic | VPA | LEV, clobazam | LEV has Class I evidence for myoclonic seizures; VPA most effective overall |
| Infantile spasms (non-TSC) | ACTH | Vigabatrin, prednisolone | ACTH superior to VGB for non-TSC spasms |
| Infantile spasms (TSC) | Vigabatrin | ACTH | VGB first-line in TSC: 65–95% response; EPISTOP: start at first EEG abnormality |
| LGS | VPA | Clobazam, rufinamide, CBD, fenfluramine | CBD (GWPCARE3/4); fenfluramine FDA-approved; callosotomy for drop attacks |
| Dravet syndrome | VPA + clobazam | Stiripentol, CBD, fenfluramine | Na+ channel blockers CONTRAINDICATED; stiripentol (STICLO: 71% vs. 5%) |
| JME | VPA (most effective) | LEV (women), LTG (caution) | LTG may worsen myoclonus; lifelong treatment required (>80–90% relapse) |
💎 Board Pearl
- Ethosuximide is first-line for CAE WITHOUT GTC; if GTC present, use VPA (ETX has no GTC efficacy)
- LTG in JME: acceptable for GTC prevention but may paradoxically worsen myoclonic jerks — use with caution
- Cenobamate: 21% seizure-free rate in phase 3 trials for drug-resistant focal epilepsy — unprecedented for adjunctive therapy
SANAD Trials
SANAD I (Lancet, 2007)
- Design: large, unblinded, randomized UK trial; first-line monotherapy comparison
- Arm A (focal): CBZ vs. GBP vs. LTG vs. OXC vs. TPM
- LTG better than CBZ for time to treatment failure; noninferior for seizure freedom
- GBP inferior to CBZ for seizure control
- Arm B (generalized/unclassified): VPA vs. LTG vs. TPM
- VPA better than TPM for treatment failure; better than LTG for seizure freedom
- Conclusion: LTG best first-line for focal; VPA best for generalized
- Limitation: unblinded design; but results have been consistent across multiple analyses
SANAD II (Lancet, 2021)
- Design: open-label, noninferiority, multicenter, phase 4 RCT
- Arm A (focal): LTG vs. LEV vs. ZNS
- LTG superior to both LEV and ZNS for time to 12-month remission
- LEV noninferior for time to treatment failure (good tolerability)
- Arm B (generalized/unclassified): VPA vs. LEV
- VPA superior to LEV for 12- and 24-month remission
- LEV noninferior for time to treatment failure
- Conclusion: confirmed LTG best for focal; VPA most effective for generalized (avoid in women of childbearing potential)
- Key takeaway for women: VPA has highest teratogenicity (>8% malformation rate); LTG or LEV preferred despite being less effective
Clinical Pearl
LTG requires slow titration (6–8 weeks) — if urgent seizure control is needed, LEV is preferred (therapeutic dose in 1–2 weeks). SANAD II confirmed LTG is slightly more effective long-term, but LEV is faster to initiate.
Seizure Freedom Probability
Glasgow Cohort (Kwan & Brodie, NEJM 2000; updated 2018)
| ASM Trial | Seizure Freedom Rate | Cumulative |
|---|---|---|
| First ASM | 47–50% | ~50% |
| Second ASM | ~13% | ~63% |
| Third ASM | ~2–5% | ~64–65% |
| Fourth+ ASMs | ~2–5% each | Marginal additional gain |
- Key message: greatest opportunity for seizure freedom is the FIRST ASM — choose wisely
- After 2 ASM failures, each subsequent drug adds only marginal benefit
- Overall: ~64% achieve seizure freedom; ~36% develop drug-resistant epilepsy
- This data underpins the ILAE 2010 definition of drug-resistant epilepsy after 2 failures
- Exception: cenobamate may achieve higher seizure-free rates (21%) even in drug-resistant populations — consider earlier in algorithm
Drug-Resistant Epilepsy
ILAE Definition (2010)
- Definition: failure of adequate trials of 2 tolerated, appropriately chosen and used ASM schedules (monotherapy or combination) to achieve sustained seizure freedom
- Adequate trial requires ALL of:
- Appropriate ASM for the seizure/epilepsy type
- Adequate dose for sufficient duration
- Tolerated (not stopped for side effects before reaching effective dose)
- Sustained seizure freedom: ≥3x longest pretreatment interseizure interval, or 12 months (whichever longer)
- Affects ~30–36% of epilepsy patients
Pseudoresistance — Rule Out Before Labeling Drug-Resistant
| Cause | Details | How to Identify |
|---|---|---|
| Nonadherence | MOST COMMON cause | Drug levels; pharmacy refill records |
| PNES | 20–30% of EMU referrals | Video-EEG; no ictal EEG correlate |
| Incorrect classification | Narrow-spectrum ASM for misclassified IGE | Reassess EEG; trial broad-spectrum ASM |
| Incorrect ASM choice | Na+ blocker worsening generalized seizures | Review medication vs. seizure type |
| Subtherapeutic dosing | Dose not optimized before switching | Serum drug levels |
| Lifestyle factors | Sleep deprivation, alcohol (especially JME) | Detailed history; seizure diary |
- After 2 ASM failures: refer for comprehensive epilepsy center evaluation
- Surgical evaluation should be offered to ALL patients with drug-resistant focal epilepsy
- Surgery achieves seizure freedom in 60–80% of selected TLE cases
- Consider cenobamate for patients not ideal surgical candidates (exceptional adjunctive efficacy)
- Neuromodulation options: VNS, RNS, DBS — FDA-approved for drug-resistant epilepsy when surgery not feasible
Synergistic ASM Combinations
Evidence-Based Rational Polytherapy
- Principle: combine different mechanisms for supra-additive efficacy and infra-additive toxicity
- Na+ channel blocker + SV2A ligand is the best-supported mechanistic pairing (e.g., LTG + LEV)
- Avoid >2 ASMs when possible — replace ineffective agents rather than stacking a third
Synergistic Pairs
| Combination | Evidence / Notes |
|---|---|
| LTG + VPA | Strongest synergy evidence. VPA inhibits LTG glucuronidation → doubles LTG levels — reduce LTG dose by 50% |
| LTG + LEV | Complementary mechanisms (Na+ channel + SV2A); no pharmacokinetic interaction |
| VPA + ethosuximide | Synergistic for refractory absence seizures |
| CBD + clobazam | CBD increases N-desmethylclobazam via CYP2C19 inhibition |
Combinations to AVOID
| Combination | Reason |
|---|---|
| BRV + LEV | Same SV2A target — receptor saturation; no additive benefit; BRV ineffective when added to LEV in trials |
| Two Na+ channel blockers (CBZ + lacosamide; CBZ + OXC) | Pharmacodynamic interaction → dizziness, diplopia, ataxia at therapeutic levels |
| Enzyme inducers + perampanel | Inducers markedly reduce perampanel levels; higher doses → more behavioral side effects |
Drugs That Worsen Specific Syndromes
Critical Exacerbation Table
| Drug(s) | Syndrome / Seizure Type Worsened | Mechanism / Notes |
|---|---|---|
| CBZ, OXC, PHT, LTG, lacosamide | Absence, myoclonic, atonic seizures in IGE | Enhance thalamocortical synchronization; may trigger NCSE |
| CBZ, OXC | Dravet syndrome | Further impair NaV1.1 in dysfunctional inhibitory interneurons; may trigger SE |
| PHT | PME (Unverricht-Lundborg) | Irreversible cerebellar atrophy — strictly contraindicated |
| VGB | LGS (except TSC spasms) | Worsens generalized epilepsy syndromes |
| GBP, PGB | Myoclonus; may worsen absence | Reported in IGE and cortical myoclonus |
| LTG | Myoclonic jerks in JME (subset) | Paradoxical worsening; may still benefit GTC prevention |
💎 Board Pearl
- Infant worsening on CBZ/OXC/PHT/LTG after febrile seizures → think Dravet, send SCN1A
- PHT in Unverricht-Lundborg (PME): irreversible cerebellar atrophy — classic board contraindication
- VGB is first-line for TSC spasms but worsens other generalized epilepsy syndromes — context is everything
ASM Withdrawal Principles
When to Consider Withdrawal
- After 2–5 years seizure-free (adults); ≥1–2 years in children
- Normal neurologic examination
- Normal or improved EEG (no new epileptiform discharges at withdrawal)
- Shared decision-making — discuss recurrence risk and driving implications
- Abnormal EEG at withdrawal is the strongest predictor of recurrence
Recurrence Risk by Syndrome
| Syndrome / Context | Recurrence Risk | Recommendation |
|---|---|---|
| Overall | ~30–40% | Individualize based on risk factors |
| JME | >80–90% | Generally lifelong treatment |
| Focal with lesion | 50–70% | Withdrawal not recommended unless post-surgical + seizure-free ≥2 years |
| CAE (normal EEG) | 20–30% | Favorable — attempt after 2 years with normal EEG |
| BECTS | <5% | Nearly all remit by mid-adolescence |
Risk Factors for Recurrence
Lower Recurrence Risk (Favors Withdrawal)
- Childhood absence epilepsy with normalized EEG
- Normal MRI, no interictal epileptiform discharges on withdrawal EEG
- Childhood-onset epilepsy; single seizure type
- Seizures controlled on first ASM (monotherapy)
- Idiopathic/genetic etiology with age-limited course
Higher Recurrence Risk (Opposes Withdrawal)
- JME (lifelong treatment paradigm)
- Epileptiform EEG at time of withdrawal — strongest predictor
- Structural brain lesion; adult onset
- Prior drug-resistant epilepsy; multiple seizure types
- Polytherapy at time of withdrawal (implies more refractory course)
Taper Protocol
- Taper gradually over 3–6 months; reduce ~25% every 2–4 weeks
- Withdraw ONE drug at a time in polytherapy; observe ≥3–6 months before withdrawing next
- BZDs/barbiturates: especially slow tapers (physical dependence; withdrawal seizure risk)
- CBZ/OXC: high rebound seizure risk with rapid withdrawal — taper ≥3 months
- Most recurrences occur within first 12 months (especially first 6 months); if seizure-free at 2 years post-withdrawal, long-term outlook favorable
- If recurrence during taper: resume previous effective dose; most patients (80–90%) regain control after reinstitution
Board Pearls
💎 Board Pearl
- SANAD I + II summary: LTG = best first-line for focal epilepsy; VPA = best for generalized. These are the highest-yield trial conclusions for board exams.
- First ASM = best chance: ~50% seizure freedom with the first ASM; drops to ~13% with the second and ~2–5% thereafter. Getting the first choice right matters most.
- Drug-resistant epilepsy = 2 ASM failures. ILAE 2010 definition. After this threshold → refer for surgical evaluation, not just another drug.
- LTG + VPA = most synergistic combination but VPA doubles LTG levels via glucuronidation inhibition. Always halve LTG dose when co-administered with VPA.
- Never combine BRV + LEV: both target SV2A; receptor saturation; no additive benefit in clinical trials.
- PHT + PME (Unverricht-Lundborg) = irreversible cerebellar atrophy. Classic drug-syndrome contraindication on boards.
- JME = lifelong treatment. >80–90% relapse after withdrawal even after decades of seizure freedom. VPA most effective; LEV in women.
Clinical Pearls
Clinical Pearl
The most common cause of apparent drug-resistant epilepsy is pseudoresistance. Always rule out nonadherence (check drug levels), PNES (20–30% of EMU referrals), and incorrect seizure classification (narrow-spectrum ASM given for misdiagnosed IGE) before labeling a patient as drug-resistant.
Clinical Pearl
Cenobamate is a game-changer for drug-resistant focal epilepsy. Phase 3 trials showed 21% seizure-free rate — unprecedented for adjunctive therapy. Requires extremely slow titration (start 12.5 mg/day) due to DRESS risk, and significantly increases clobazam levels.
Clinical Pearl
LTG + VPA polytherapy withdrawal: if VPA is withdrawn first, LTG levels DROP (VPA was inhibiting LTG clearance). If LTG is withdrawn first, no pharmacokinetic change in VPA. Anticipate this interaction to avoid breakthrough seizures.
References
- Marson AG, Al-Kharusi AM, Alwaidh M, et al. The SANAD study of effectiveness of carbamazepine, gabapentin, lamotrigine, oxcarbazepine, or topiramate for treatment of partial epilepsy. Lancet 2007;369(9566):1000–1015.
- Marson AG, Al-Kharusi AM, Alwaidh M, et al. The SANAD study of effectiveness of valproate, lamotrigine, or topiramate for generalised and unclassifiable epilepsy. Lancet 2007;369(9566):1016–1026.
- Marson A, Burnside G, Appleton R, et al. The SANAD II study of the effectiveness and cost-effectiveness of levetiracetam, zonisamide, or lamotrigine for newly diagnosed focal epilepsy. Lancet 2021;397(10282):1363–1374.
- Marson A, Burnside G, Appleton R, et al. The SANAD II study of the effectiveness and cost-effectiveness of valproate versus levetiracetam for newly diagnosed generalised and unclassifiable epilepsy. Lancet 2021;397(10282):1375–1386.
- Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med 2000;342(5):314–319.
- Chen Z, Brodie MJ, Liew D, Kwan P. Treatment outcomes in patients with newly diagnosed epilepsy treated with established and new antiepileptic drugs: a 30-year longitudinal cohort study. JAMA Neurol 2018;75(3):279–286.
- Kwan P, Arzimanoglou A, Berg AT, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 2010;51(6):1069–1077.
- Glauser TA, Cnaan A, Shinnar S, et al. Ethosuximide, valproic acid, and lamotrigine in childhood absence epilepsy. N Engl J Med 2010;362(9):790–799.
- Abou-Khalil B. Update on antiseizure medications 2025. Continuum (Minneap Minn) 2025;31(1):123–165.
- Krauss GL, Klein P, Brandt C, et al. Safety and efficacy of adjunctive cenobamate (YKP3089) in patients with uncontrolled focal seizures. Lancet Neurol 2020;19(1):38–48.