When to Stop Antiseizure Medication: A Practical Guide for Seizure-Free Patients
Clinical Practice Update — Assessing Recurrence Risk, Shared Decision-Making, Taper Strategies, and Managing Consequences of Withdrawal
This is an original clinical education article informed by current guidelines and evidence. See References below for source documents.
- Clinical Focus
- Antiseizure medication withdrawal in seizure-free adults and children, recurrence risk estimation, prognostic factors, taper protocols, driving implications, and management of relapse
- Target Audience
- Neurologists (adult and paediatric), epileptologists, primary care physicians, pharmacists, nurse practitioners, neurology residents
- Setting
- Outpatient neurology, paediatric neurology, primary care, post-surgical epilepsy follow-up
- Source Evidence
- •AAN Practice Advisory Update — ASM Withdrawal in Seizure-Free Patients (Gloss et al., Neurology 2021; Reaffirmed 2025)
- •MRC Antiepileptic Drug Withdrawal Study (Lancet 1991; re-analysis Terman et al., Epilepsia 2022)
- •Individual Participant Data Meta-Analysis — Prediction Model for Withdrawal Outcomes (Lamberink et al., Lancet Neurol 2017)
- •Pooled Analysis — Variation in Seizure Risk From ASM Withdrawal (Terman et al., Epilepsia Open 2024)
- •Akershus Study — Double-Blind RCT of AED Withdrawal in Adults (Lossius et al., Epilepsia 2008)
Key Clinical Takeaways
The most important actionable points from this Practice Update on antiseizure medication withdrawal. Each links to the detailed discussion below.
- 1ASM withdrawal can be considered after at least 2 years of seizure freedom in adults, and 1.5–2 years in children — but the decision is always individualised → When to Consider
- 2Overall seizure recurrence after ASM withdrawal is approximately 30–50%, with the highest risk in the first 12 months → Recurrence Risk
- 3Shared decision-making is mandatory — explore the patient's quality of life, concerns, driving status, and personal tolerance for seizure risk → Shared Decision-Making
- 4Key predictors of relapse include longer epilepsy duration before remission, epileptiform EEG before withdrawal, structural aetiology, and absence of a self-limiting syndrome → Risk Factors
- 5The role of EEG and imaging before withdrawal remains uncertain — the AAN states the evidence is insufficient to confirm their predictive value → Role of Investigations
- 6If seizures recur after withdrawal, the majority of patients regain seizure control when treatment is restarted — but up to 20% do not achieve immediate remission → Managing Relapse
- 7Taper gradually over weeks to months — there is no strong evidence to favour a specific taper schedule → How to Taper
- 8Driving restrictions apply during and after withdrawal — counsel patients about local regulations before starting the taper → Driving and Lifestyle
- 9JME and structural focal epilepsies carry a high relapse risk — withdrawal should be approached with particular caution → High-Risk Syndromes
- 10A validated online prediction tool exists to estimate individualised recurrence risk — consider using it in patient consultations → Prediction Tools
When Should You Consider Stopping ASM Treatment?
The majority of people started on antiseizure medication for epilepsy — roughly 60–70% — will eventually become seizure-free. Once that milestone is reached, a difficult question follows: is lifelong treatment necessary, or can the medication be safely stopped? Few decisions in neurology provoke as much anxiety in both patients and clinicians. On one side sits the real burden of ongoing treatment — adverse effects, cost, teratogenicity risk, and the psychosocial weight of being "on medication for epilepsy." On the other sits the possibility of seizure recurrence and everything that comes with it: injury, loss of driving privileges, employment consequences, and psychological impact.
Confirm the diagnosis before anything else. Reassess the original diagnosis before initiating withdrawal. Ensure the patient genuinely met diagnostic criteria for epilepsy and is not on prophylactic treatment for a condition that no longer warrants it (e.g., acute symptomatic seizures, resolved structural lesion).
Strong Rec Low Evidence Expert ConsensusQuantify the recurrence risk for adults. Inform seizure-free adults that stopping ASMs after a minimum of 2 years without seizures roughly doubles the chance of recurrence relative to staying on treatment. In absolute terms, the 2-year recurrence rate sits around 40–45% after withdrawal versus approximately 20% with continued therapy.
Strong Rec High Evidence AAN 2021 MRC 1991 Terman 2024Paediatric timing may be shorter. For seizure-free children and adolescents, the available evidence suggests that beginning withdrawal after 1.5–2 years of seizure freedom does not substantially change the recurrence rate compared with waiting 4 years. Discuss this with families to set expectations.
Moderate Rec Moderate Evidence AAN 2021Centre the patient's priorities. Clinicians should explore the factors that matter most to the individual patient when discussing ASM discontinuation — including adverse effects, driving eligibility, employment, pregnancy planning, teratogenicity concerns, and the patient's personal tolerance for seizure risk. The goal is a genuinely shared decision, not a clinician-directed recommendation.
Strong Rec Moderate Evidence AAN 2021 (Level A)What Predicts Seizure Recurrence After ASM Withdrawal?
A large individual participant data meta-analysis pooling data from 1,769 patients identified several independent predictors of seizure recurrence. More recent pooled analyses confirm that the absolute increase in 2-year seizure risk from withdrawal ranges from roughly 7% in the lowest-risk patients to 37% in the highest-risk patients, with a median increase around 19%.
Advise patients with juvenile myoclonic epilepsy (JME) or focal epilepsy with a structural aetiology that these syndromes carry a particularly high risk of relapse after withdrawal, even after many years of seizure freedom. Exercise caution and involve the patient fully in the decision.
Strong Rec Moderate Evidence AAN 2021 Lamberink 2017Consider using a validated online prediction tool (e.g., the Lamberink epilepsy prediction nomogram) to estimate individualised 2-year and 5-year seizure recurrence risk, which incorporates multiple prognostic factors simultaneously.
Moderate Rec Moderate Evidence Lamberink 2017Note that epileptiform discharges on EEG before withdrawal are associated with a higher risk of seizure recurrence, though the evidence remains insufficient to confirm that EEG or imaging definitively changes the withdrawal decision. If obtained, interpret results as one data point among many — not as a go/no-go signal.
Moderate Rec Low Evidence AAN 2021 (Level B) Lamberink 2017Factors That Influence Seizure Recurrence Risk After ASM Withdrawal
| Prognostic Factor | Direction of Effect | Strength of Evidence | What It Means in Practice |
|---|---|---|---|
| EPILEPSY CHARACTERISTICS | |||
| JME or structural focal epilepsy | Higher risk | Strong | Most experts recommend lifelong treatment for JME; surgical candidates may differ |
| Self-limiting epilepsy syndrome (e.g., SeLECTS, CAE) | Lower risk | Strong | Withdrawal is often straightforward in these patients; relapse rate is low |
| Higher number of seizures before remission | Higher risk | Strong | A heavy seizure burden before control suggests a more refractory course |
| TREATMENT HISTORY | |||
| Polytherapy before withdrawal | Higher risk | Moderate | Needing multiple ASMs to achieve control implies the epilepsy is harder to suppress |
| Longer epilepsy duration before remission | Higher risk | Strong (IPD meta-analysis) | Patients who took years to become seizure-free have a more treatment-resistant phenotype |
| Shorter seizure-free interval before withdrawal | Higher risk | Strong | Waiting longer before withdrawal reduces risk; 2 years is the minimum |
| INVESTIGATIONS | |||
| Epileptiform EEG before withdrawal | Higher risk | Moderate (consistent across studies, but AAN notes uncertainty) | Consider obtaining EEG before withdrawal; interpret with caution |
| PATIENT-CENTRED FACTORS (not captured in prediction models) | |||
| Driving status and lifestyle consequences | Does not change biological risk, but strongly affects the decision | Survey data | Patients who rely on driving or have high-risk occupations often decline withdrawal, regardless of recurrence probability |
| Medication burden and adverse effects | Does not change biological risk, but may tip the decision towards withdrawal | Expert consensus | Significant cognitive, metabolic, or teratogenic adverse effects create a stronger case for attempting withdrawal, even at moderate recurrence risk |
- No single factor should be used in isolation to decide for or against withdrawal. The value of prediction tools is that they combine multiple factors into an individualised risk estimate.
Clinical Decision Pathway
A practical, question-based approach to the ASM withdrawal decision.
How Should You Taper, and What Should You Monitor?
Set expectations about what happens if seizures return. Inform patients that if seizures recur after withdrawal, most people regain seizure control when the previously effective ASM is restarted. However, up to 20% may not achieve immediate remission, and in rare cases, the restarted medication may be less effective than before. This information is essential before the taper begins, not after relapse.
Strong Rec Moderate Evidence AAN 2021Warn about the highest-risk window. The first 6–12 months after completing withdrawal carry the greatest recurrence risk. Arrange follow-up during this window and ensure the patient knows to seek urgent review if a seizure occurs.
Strong Rec High Evidence MRC 1991 Lamberink 2017Taper gradually. Reduce doses over at least 2–3 months for most ASMs. No strong evidence supports one taper schedule over another, but abrupt discontinuation should be avoided.
Moderate Rec Low Evidence AAN 2021Exercise special caution with barbiturates and benzodiazepines. These drugs require a longer taper (3–6 months or more) to avoid withdrawal seizures — a pharmacological phenomenon distinct from epilepsy relapse.
Strong Rec Moderate Evidence Expert ConsensusPractical Monitoring During and After ASM Withdrawal
| Phase | What to Do | Key Considerations | Common Pitfalls |
|---|---|---|---|
| Before starting taper | Confirm diagnosis, discuss driving rules, consider EEG, use prediction tool | Ensure the patient fully understands the recurrence risk and consequences | Forgetting to discuss driving restrictions — patients lose their licence mid-taper and feel blindsided |
| During taper | Reduce dose gradually over 2–3+ months; check in at each dose step | Barbiturates/benzodiazepines need slower taper (3–6 months) | Tapering too fast; not monitoring for new seizures during dose reduction |
| First 6–12 months post-withdrawal | Highest risk period; follow up at 3 and 6 months minimum | Ensure patient can contact clinic urgently if a seizure occurs | Discharging the patient too soon after stopping the ASM |
| If seizure recurs | Restart the previously effective ASM promptly; reassess driving status | Most patients regain control; up to 20% may not achieve immediate remission | Delaying ASM restart; not recounselling about long-term treatment expectations |
Evidence in Context
The MRC Trial: What Changed Our Understanding
The 1991 MRC Antiepileptic Drug Withdrawal Study remains the largest randomised trial of ASM withdrawal, enrolling over 1,000 seizure-free patients across 40 UK centres. It showed that withdrawal approximately doubled the risk of seizure recurrence compared with continuing treatment. A recent re-analysis using contamination-adjusted methods found the actual effect may be slightly larger than originally reported — with a 2-year recurrence risk of roughly 43% for withdrawal versus 21% for continuation when accounting for crossover between arms. This trial also demonstrated that psychiatric factors and shorter seizure-free periods were independent predictors of recurrence.
The Prediction Nomogram: Making Risk Personal
The Lamberink et al. (2017) individual participant data meta-analysis pooled data from 1,769 patients to create evidence-based nomograms that calculate individualised 2-year and 5-year seizure recurrence risk after ASM withdrawal. The model incorporates epilepsy duration before remission, seizure-free interval, number of seizures before remission, age at onset, history of febrile seizures, self-limiting syndrome status, developmental delay, and EEG before withdrawal. The discrimination was moderate (concordance statistic ~0.65–0.71) but stable across populations. This tool has been made available online and is increasingly used in clinical consultations to move beyond generic recurrence rates toward personalised counselling.
Post-Surgical Patients: A Special Case
The AAN practice advisory notes that the evidence regarding ASM withdrawal after epilepsy surgery is particularly weak, with only one low-quality trial available. The decision to withdraw ASM in post-surgical patients who are seizure-free involves unique considerations: the resection may have removed the epileptogenic focus entirely, but the risk of recurrence from residual or new foci remains. Most epilepsy surgery centres recommend at least 1–2 years of seizure freedom post-surgery before considering withdrawal, with decisions guided by the completeness of resection, post-operative EEG, and MRI findings. This remains an area of active research.
Why Patients and Clinicians Often Disagree
A recent cross-sectional study found a striking disconnect: neurologists recommended discontinuation in 56% of eligible seizure-free patients, but only 15% of patients ultimately agreed to complete withdrawal. Longer seizure-free duration paradoxically made patients more reluctant to stop (they had more to lose), while it made neurologists more willing to recommend it. Patients with a history of generalised tonic-clonic seizures, higher anxiety scores, and those who tolerated their ASM well were particularly reluctant. Understanding these differing perspectives is critical to genuine shared decision-making.
What We Still Don't Know
References
- 1.Gloss D, Pargeon K, Pack A, et al. Antiseizure Medication Withdrawal in Seizure-Free Patients: Practice Advisory Update Summary. Neurology. 2021;97(23):1072–1081. doi:10.1212/WNL.0000000000012944
- 2.MRC Antiepileptic Drug Withdrawal Study Group. Randomised study of antiepileptic drug withdrawal in patients in remission. Lancet. 1991;337(8751):1175–1180. doi:10.1016/0140-6736(91)92856-W
- 3.Lamberink HJ, Otte WM, Geerts AT, et al. Individualised prediction model of seizure recurrence and long-term outcomes after withdrawal of antiepileptic drugs in seizure-free patients. Lancet Neurol. 2017;16(7):523–531. doi:10.1016/S1474-4422(17)30114-X
- 4.Terman SW, Slinger G, Koek A, et al. Variation in seizure risk increases from antiseizure medication withdrawal among patients with well-controlled epilepsy: a pooled analysis. Epilepsia Open. 2024;9(1):333–344. doi:10.1002/epi4.12880
- 5.Lossius MI, Hessen E, Mowinckel P, et al. Consequences of antiepileptic drug withdrawal: a randomized, double-blind study (Akershus Study). Epilepsia. 2008;49(3):455–463. doi:10.1111/j.1528-1167.2007.01323.x
- 6.Terman SW, Wang C, Wang L, et al. Reappraisal of the Medical Research Council Antiepileptic Drug Withdrawal Study: contamination-adjusted and dose-response re-analysis. Epilepsia. 2022;63(7):1724–1735. doi:10.1111/epi.17273
How to Read the Evidence Tags
Every recommendation carries two tags indicating recommendation strength and evidence quality. These are Medaptly's own simplified interpretations for educational clarity.
Recommendation Strength
| Tag | What It Means | In Practice |
|---|---|---|
| Strong Rec | Broadly supported by evidence or expert consensus. Benefits clearly outweigh risks. | Standard practice for most patients meeting the criteria. |
| Moderate Rec | Evidence favours this action, with some uncertainty. | Appropriate for most; clinical context may guide alternatives. |
| Conditional Rec | Benefit less certain; depends on individual circumstances. | Use shared decision-making. |
| Against | Evidence shows no benefit or potential harm. | Avoid. Document reasoning if used in unusual circumstances. |
Evidence Quality
| Tag | What It Means | Confidence Level |
|---|---|---|
| High Evidence | Multiple RCTs, large meta-analyses, or well-validated models. | Very confident. Unlikely to change substantially. |
| Moderate Evidence | Single RCT, large observational studies, or IPD meta-analyses. | Reasonably confident. Direction likely correct. |
| Low Evidence | Expert consensus, small studies, or conflicting data. | Less certain. Best available guidance; may evolve. |
These are Medaptly's simplified interpretations for educational clarity. For the full classification systems used by each source, consult the original documents listed in References.