AAN/AES Guideline: Management of an Unprovoked First Seizure in Adults (2015)

This topic summarizes the 2015 AAN/AES evidence-based guideline on the prognosis and treatment of an unprovoked first seizure in adults, published in Neurology 2015;84:1705–1713. The guideline reviewed 47 articles addressing recurrence risk, treatment effects, and adverse events.

Guideline Overview

Source & Scope

• Organizations: AAN Guideline Development Subcommittee + American Epilepsy Society (AES)
• Published: Neurology 2015;84:1705–1713 (April 21, 2015)
• Authors: Krumholz A, Wiebe S, Gronseth GS, Gloss DS, Sanchez AM, Kabir AA, Liferidge AT, Martello JP, Kanner AM, Shinnar S, Hopp JL, French JA
• Companion guideline: 2007 AAN guideline addressed evaluation of first seizure; this 2015 guideline addresses prognosis and treatment
• Search: MEDLINE, Embase, Cochrane Central Register (1966–March 2013) → 2,613 articles identified → 281 full-text → 47 accepted and rated
• Population: Adults with an unprovoked first seizure; excluded patients with >1 seizure at presentation

Key Definitions

• Unprovoked seizure: Seizure of unknown etiology or seizure related to a demonstrated preexisting brain lesion/progressive CNS disorder (“remote symptomatic”)
• Provoked seizure (excluded): Seizure due to an acute symptomatic condition (metabolic, toxic, cerebral trauma, stroke)
• Immediate treatment: AED started within 1 week of index seizure (Class I study); within 1 month in 55%, within 3 months in 81% (Class II study)
• Deferred treatment: AED withheld until a second seizure occurs

Three Clinical Questions

1. What are the risks for seizure recurrence after a first seizure?
2. Does immediate AED treatment change short-term risk for recurrence or long-term prognosis for seizure remission?
3. For those prescribed AEDs immediately, what are the risks for adverse events?

AAN Evidence Classification Scheme (for this guideline)

Q1: Risk of Seizure Recurrence

Overall Recurrence Rates (Pooled Data, Mixed Treated & Untreated)

• Key pattern: Most recurrences occur within the first 1–2 years, especially in year 1
• Recurrence risk is 21%–45% within 2 years (Level A)
• Cumulative risk continues to rise slowly after 2 years but the rate of new recurrences diminishes
• Total pooled cohort: 3,212 patients across 10 studies (2 Class I, 8 Class II)
• Seizure recurrence was lower in AED-treated patients, but most studies did not randomize treatment

Risk Factors That INCREASE Recurrence

Recurrence in Remote Symptomatic vs Unknown Cause

• Remote symptomatic seizure: ~2-fold higher recurrence risk
• Representative recurrence rates (from Class I study):
  • Remote symptomatic: 26% at 1 yr, 41% at 3 yr, 48% at 5 yr
  • Unknown cause: 10% at 1 yr, 24% at 3 yr, 29% at 5 yr
• ~10% of subjects had clinically relevant structural lesions on imaging

Factors NOT Consistently Associated with Recurrence

• Patient age
• Sex
• Family history of seizures
• Seizure type (focal vs generalized)
• Presentation with status epilepticus
• Multiple (≥2) discrete seizures within 24 hours with recovery between them

Q2a: Immediate AED Treatment — Short-Term Effect on Recurrence (2 Years)

Evidence Summary

• 1 Class I study (FIRST trial) + 4 Class II studies (including MESS trial)
• Immediate AED therapy significantly reduces seizure recurrence within 2 years
• Absolute risk reduction: ~35% (95% CI 23%–46%) from random-effects meta-analysis
• Class I study: “immediate” = AED started within 1 week of index seizure
• Class II study: within 1 week in 30%, by 1 month in 55%, by 3 months in 81%

Short-Term Recurrence: Immediate vs Deferred Treatment (Table 2 Data)

* Statistically significant difference (p < 0.05)

• The Class I study (FIRST) and 3 of 4 Class II studies showed significantly fewer recurrences with immediate AED treatment
• MESS trial (largest, Class II): smaller magnitude of difference (32% vs 39%), but MESS included patients with multiple seizures before randomization

Quality of Life

• Only 1 Class II study (MESS/Jacoby) assessed QOL
• No significant difference in standard, validated 2-year QOL measures between immediate and deferred treatment groups
• Patients NOT immediately treated were more likely to be restricted from driving

Conclusion (Level B)

• Immediate AED therapy reduces absolute recurrence risk by ~35% within 2 years
• Immediate AED therapy may not improve QOL (Level C)

Q2b: Immediate AED Treatment — Long-Term Effect on Seizure Remission (>3 Years)

Evidence Summary

• 1 Class I study (FIRST trial) + 1 Class II study (MESS trial)
• Long-term outcome measured by sustained seizure remission (seizure freedom for a specified duration, typically 2–5 years)

Seizure Remission: Immediate vs Deferred Treatment (Table 3 Data)

NS = not a significant difference. * 5-year remission rates also not significantly different. ** 5-year seizure remission in patients followed longer was also not significantly different.

• No difference in 2-year sustained seizure remission between immediate and deferred treatment (both studies)
• 5-year seizure remission: also not significantly different
• Mortality: 1 Class III study with 20-year follow-up found immediate treatment does not affect mortality

Conclusion (Level B)

• Over the longer term (>3 years), immediate AED treatment is unlikely to improve the chance of attaining sustained seizure remission
• Whether you treat now or wait until a second seizure, the long-term prognosis is the same

Q3: Risks of AED Adverse Events

Evidence Summary

• 4 Class II studies + 1 Class III study
• AEDs studied were primarily older drugs: phenytoin, phenobarbital, carbamazepine, valproic acid, lamotrigine
• Studies focused on patients with an unprovoked first seizure immediately treated with a single AED

Adverse Event Rates

• AEs are typically dose-related and reversible through dose reduction or drug switch
• AEDs used in these studies are now considered older-generation agents; newer AEDs may have fewer and different AEs
• One Class III study found AE-related AED discontinuation in only 7%–13%, because drugs were started as monotherapy at low doses

Clinical Context & Decision-Making

Recurrence Risk Stratification

• Recurrence can be estimated and stratified by clinical risk factors:
  • Prior brain insult/lesion (Level A)
  • Epileptiform EEG (Level A)
  • Abnormal brain imaging (Level B)
  • Nocturnal seizure (Level B)
• In some patients, the statistical recurrence risk may approach that of patients with multiple seizures (i.e., established epilepsy)
• Risk of recurrence is very high after additional seizures: 57% by 1 year and 73% by 4 years after a second unprovoked seizure

ILAE 2014 Definition of Epilepsy

• The ILAE expanded the definition of epilepsy beyond the prior standard of ≥2 unprovoked seizures
• Now includes: 1 unprovoked seizure + a high (≥60%) recurrence risk over 10 years
• However, the guideline cautions: evidence on specific risk factors and their interactions is lacking
• “No formula can be applied for additive risks… such risks will have to be decided by individualized considerations.” — ILAE report

When to Treat

• AED treatment is generally accepted when a patient has ≥2 unprovoked seizures (“epilepsy”)
• After a first seizure: treatment is debated and must be individualized
• Decision should weigh:
  • Individual recurrence risk (based on risk factors above)
  • AED adverse event profile
  • Patient preferences & values
  • Social consequences of recurrence (driving, employment, safety)
  • The knowledge that immediate AEDs will NOT improve long-term remission

Driving Implications

• Only 21% of patients with first seizures received correct advice about driving limitations
• Patients not immediately treated with AEDs were more likely to be restricted from driving (MESS QOL study)
• Driving laws vary by jurisdiction — individual state requirements should be reviewed

Caution on Additive Risks

• Only 2 studies analyzed additive effects or covariance of risk factors for seizure recurrence
• One study: only independent risk factor was epileptiform EEG
• The other study: only independent risk factor was remote symptomatic etiology
• Cannot simply add risk factors together — interactions are poorly understood

Key Trials Referenced

FIRST Trial (First Seizure Trial Group)

• Design: Randomized controlled trial — Class I evidence
• Population: 397–419 adults with first unprovoked tonic-clonic seizure
• Intervention: Immediate AED treatment within 1 week vs deferred until recurrence
• Short-term result: Immediate treatment: 18% recurrence vs deferred: 39% (p < 0.05) at 2 years
• Long-term result: 2-year sustained remission: 81% immediate vs 78% deferred (NS); 5-year remission also NS
• 20-year mortality: No difference (Class III study, Leone et al. 2006)
• Conclusion: Immediate AEDs reduce short-term recurrence but do NOT improve long-term seizure freedom or mortality

MESS Trial (Marson et al., Lancet 2005)

• Design: Randomized controlled trial — Class II evidence
• Population: 812 patients (included those with single and multiple seizures before randomization)
• Intervention: Immediate vs deferred AED treatment
• Short-term result: Immediate: 32% recurrence vs deferred: 39% at 2 years
• Long-term result: 5-year seizure remission: 92% in both groups (NS)
• QOL: No significant difference in standard 2-year validated QOL measures
• Driving: Non-treated patients more likely to be restricted from driving

Seizure Recurrence Data by Study (Table 1)

Risk of Recurrence at Various Time Points (Class I & II Studies, N = 3,212)

• Table based on a fixed-effect pooled percentage model
• GTC seizures comprised the major seizure type across studies
• Wide variation in recurrence rates reflects differences in patient ascertainment, treatment, and follow-up duration

Summary of Recommendations by Evidence Level

Quick Reference: Evidence Levels at a Glance

Future Research Needs

• Studies on patient management techniques, interventions, and counseling focusing on objective outcomes (QOL) are needed
• How, when, and by whom patients should receive driving advice
• Further studies on patient preferences, psychosocial factors, and QOL measures
• Updated studies using newer AEDs for initial treatment (existing data based on older AEDs)
• Research on AED discontinuation in patients with a first seizure who receive AEDs
• Predictive statistical models to analyze additive recurrence risks associated with specific clinical variables, EEG findings, and brain imaging
• Better data on the degree to which AED treatment may influence recurrence risk for each clinical risk factor individually and in combination