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.
🔹 Bottom Line
- Guideline: AAN/AES 2015 evidence-based guideline — 47 articles reviewed (2 prognostic Class I, 8 Class II; 1 therapeutic Class I, 4 Class II) addressing recurrence risk & treatment
- Recurrence risk: 21%–45% within the first 2 years (Level A); greatest in year 1 (~32% at 1 yr, ~46% by 5 yr)
- Risk factors for recurrence: Prior brain insult (Level A), epileptiform EEG (Level A), significant brain-imaging abnormality (Level B), nocturnal seizure (Level B)
- Immediate AED therapy reduces seizure recurrence risk by ~35% absolute over 2 years (Level B), but does NOT improve long-term prognosis for sustained seizure remission (Level B)
- Quality of life: Immediate AED treatment may NOT improve QOL (Level C)
- AED adverse events: Occur in 7%–31% of patients (Level B); predominantly mild and reversible
- Treatment decision: Individualized — weigh recurrence risk vs AED side effects; educate patient that AEDs reduce 2-year risk but do NOT change long-term remission
- NOT consistently associated with recurrence: Age, sex, family history, seizure type, status epilepticus, multiple seizures within 24 hours
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
- What are the risks for seizure recurrence after a first seizure?
- Does immediate AED treatment change short-term risk for recurrence or long-term prognosis for seizure remission?
- For those prescribed AEDs immediately, what are the risks for adverse events?
AAN Evidence Classification Scheme (for this guideline)
| Level | Evidence Required | Recommendation Strength |
|---|
| A | ≥2 consistent Class I studies or 1 Class I + ≥2 Class II | Established — should be done |
| B | ≥1 Class I or ≥2 consistent Class II | Probable — should be considered |
| C | ≥1 Class II or ≥2 consistent Class III | Possible — may be considered |
| U | Insufficient or conflicting evidence | Data inadequate |
Q1: Risk of Seizure Recurrence
Overall Recurrence Rates (Pooled Data, Mixed Treated & Untreated)
| Time After First Seizure | Cumulative Recurrence (%) | Studies |
|---|
| 1 month | 7 (64 of 1,196 treated; total 220 of 3,212) | 10 Class I & II |
| 3 months | 18 (519 of 3,212) | Pooled |
| 6 months | 24 (761 of 3,212) | Pooled |
| 1 year | 32 (873 of 3,212) | Pooled |
| 2 years | 36 (508 of 3,212) | Pooled |
| 3 years | 42 | Pooled |
| 5 years | 46 (685 of 3,212) | Pooled |
| >5 years | 49 (723 of 3,212) | Pooled |
- 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
| Risk Factor | Relative Risk / OR (95% CI) | Timeframe | Evidence |
|---|
| Prior brain insult (stroke, trauma, CNS infection, CP, cognitive disability — “remote symptomatic”) |
RR 2.55 (1.44–4.51) |
1–5 years |
2 Class I, 2 Class II → Level A |
| EEG with epileptiform abnormalities (spikes or sharp waves) |
RR 2.16 (1.07–4.38) |
1–5 years |
2 Class I, 4 Class II → Level A |
| Significant brain-imaging abnormality (MRI/CT showing structural lesion judged as seizure cause) |
HR 2.44 (1.09–5.44) |
1–4 years |
2 Class II, 1 Class III → Level B |
| Nocturnal seizure |
OR 2.1 (1.0–4.3) |
1–4 years |
2 Class II → Level B |
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
🔹 Clinical Pearl
The four established risk factors for recurrence after a first unprovoked seizure are: (1) prior brain insult, (2) epileptiform EEG, (3) abnormal brain imaging, and (4) nocturnal seizure. Family history, sex, age, and seizure type do NOT reliably predict recurrence. A “remote symptomatic” seizure (prior stroke, TBI, etc.) roughly doubles recurrence risk compared to seizures of unknown cause.
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)
| Study (Ref.) | Class | N | Treated, n (%) | Recurrence Treated, n (%) | Recurrence Untreated, n (%) | Follow-up (yr) |
|---|
| FIRST (12–14) | I | 397 | 204 (51) | 36 (18)* | 75 (39) | 2 |
| Ref. 18 | II | 76 | 36 (47) | 4 (11)* | 18 (45) | 1 |
| MESS (15) | II | 812 | 404 (50) | 129 (32) | 159 (39) | 2 |
| Ref. 21 | II | 228 | 113 (50) | 5 (4)* | 63 (55) | 1 |
| Ref. 22 | II | 87 | 45 (52) | 9 (20)* | 28 (66) | 2 |
| Total | — | 1,600 | 804 (50) | 183 (23) | 343 (43) | 1–2 |
* 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)
🔹 Clinical Pearl
Immediate AED treatment after a first seizure reduces 2-year recurrence risk by about 35% absolute (NNT ~3), but this benefit does NOT translate into improved quality of life. The QOL finding is a common board question — treating a first seizure helps prevent a second seizure in the short term, but does not make the patient “feel better” overall.
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)
| Study (Ref.) | Class | N | Immediate Rx, n (%) | Remission Immediate, n (%) | Remission Deferred, n (%) | Follow-up |
|---|
| FIRST (12–14) | I | 419 | 215 (51) | 174 (81), NS | 159 (78) | >3 years* |
| MESS (15) | II | 812 | 404 (50) | 372 (92), NS | 375 (92) | 5 years** |
| Total | — | 1,231 | 619 (50) | 546 (88) | 534 (87) | — |
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
🔹 Clinical Pearl
Immediate AED treatment after a first seizure does NOT change the long-term prognosis. Whether started immediately or deferred until a second seizure, patients achieve the same rate of sustained seizure remission at 3–5 years. This is one of the most important takeaways for board exams — AEDs after a first seizure reduce the short-term risk of a second seizure but do not alter the natural history of epilepsy.
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
| Finding | Data | Evidence Level |
|---|
| Overall AE incidence | 7%–31% across a variety of AEDs | Level B (4 Class II, 1 Class III) |
| Severity | Predominantly mild and reversible | Level B |
| AED-related deaths | None reported | — |
| Life-threatening allergic reactions | None reported | — |
| AED discontinuation due to AE | 7%–13% (PHT or topiramate monotherapy) | Class III |
| AEs more vs less likely than controls | No more likely in epilepsy patients than untreated controls in one study | Class II |
- 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 Pearl
AED adverse events after a first seizure range from 7% to 31% but are predominantly mild and reversible. No deaths or life-threatening allergic reactions were reported. These data are based on older AEDs — newer agents may have lower AE rates. The guideline emphasizes that AE risk should be weighed against the individual patient’s recurrence risk in a shared decision-making model.
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
🔹 Clinical Pearl
A common board scenario: Patient has a first unprovoked seizure with a normal EEG, normal MRI, and daytime occurrence. What do you advise? Answer: The 2-year recurrence risk is ~21%–45% overall but lower in this low-risk patient. Immediate AED treatment would reduce short-term recurrence but NOT improve long-term remission or QOL. A shared decision-making approach is recommended (Level B). Conversely, if the patient has a prior stroke + epileptiform EEG + nocturnal seizure, the risk approaches that of established epilepsy, and treatment is more strongly warranted.
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)
| Ref. | Class | Age | N | Treated, n (%) | 1 mo (%) | 3 mo (%) | 6 mo (%) | 1 yr (%) | 2 yr (%) | 3 yr (%) | 5 yr (%) | >5 yr (%) |
|---|
| 10, 11 | I | 70% >19 | 238 | 164 (69) | — | — | — | 38 (16) | 50 (21) | 60 (29) | 70 (34) | 81 (39) |
| 12, 13 | I | 72% >16 | 397 | 204 (51) | 24 (6) | 58 (15) | 75 (19) | 98 (25) | 111 (28) | — | — | — |
| 17 | II | ≥16 | 147 | 62 (42) | — | — | 39 (27) | 50 (34) | 60 (41) | 61 (41) | — | — |
| 18 | II | Mean >20 | 76 | 36 (47) | 2 (3) | 18 (24) | 20 (26) | 22 (29) | — | — | — | — |
| 16 | II | ≥16 | 306 | 41 (13) | — | 55 (18) | 79 (26) | 111 (36) | 136 (44) | 144 (47) | — | — |
| 19 | II | 75% >15 | 424 | ? | 38 (9) | 89 (21) | 127 (30) | 153 (36) | 191 (45) | 204 (48) | 237 (56) | 244 (58) |
| 20 | II | 14–91 | 497 | 127 (26) | — | — | — | 191 (38) | — | — | — | — |
| 15 | II | 60% >20 | 812 | 404 (50) | — | 179 (22) | — | 288 (35) | — | 378 (46) | 398 (49) | — |
| 21 | II | ≥16 | 228 | 113 (50) | — | — | — | 68 (30) | — | — | — | — |
| 22 | II | 18–50 | 87 | 45 (52) | — | — | — | 30 (34) | 37 (43) | 39 (45) | — | — |
| Total | — | — | 3,212 | 1,196 (43) | 64 (7) | 220 (18) | 519 (24) | 761 (32) | 873 (36) | 508 (42) | 685 (46) | 723 (49) |
- 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
| # | Recommendation | Level |
|---|
| 1 |
Adults with an unprovoked first seizure should be informed that the chance of recurrence is greatest within the first 2 years (21%–45%) |
A |
| 2 |
Clinicians should advise that clinical factors associated with increased recurrence risk include a prior brain insult (e.g., stroke, trauma) |
A |
| 3 |
Clinicians should advise that an EEG with epileptiform abnormalities is associated with increased recurrence risk |
A |
| 4 |
Clinicians should consider advising that a significant brain-imaging abnormality is associated with increased recurrence risk |
B |
| 5 |
Clinicians should consider advising that a nocturnal seizure is associated with increased recurrence risk |
B |
| 6 |
Clinicians should consider advising that immediate AED therapy reduces recurrence risk in the 2 years subsequent to a first seizure |
B |
| 7 |
Clinicians may advise that immediate AED therapy may not improve QOL |
C |
| 8 |
Clinicians should consider advising that over the longer term (>3 years), immediate AED treatment is unlikely to improve the prognosis for sustained seizure remission |
B |
| 9 |
Patients should be advised that risk for AED adverse events ranges from 7% to 31% and these AEs are predominantly mild and reversible |
B |
| 10 |
Treatment recommendations should be based on individualized assessments weighing recurrence risk against AED AEs, educated patient preferences, and the understanding that immediate treatment reduces 2-year seizure risk but does NOT improve long-term prognosis |
Practice recommendation (consensus) |
Quick Reference: Evidence Levels at a Glance
| Level A (Established) | Level B (Probable) | Level C (Possible) |
|---|
- Recurrence risk 21%–45% in 2 yr
- Prior brain insult ↑ risk
- Epileptiform EEG ↑ risk
|
- Abnormal imaging ↑ risk
- Nocturnal seizure ↑ risk
- Immediate AED ↓ 2-yr recurrence
- Immediate AED does NOT improve long-term remission
- AED AEs: 7%–31%, mild & reversible
|
- Immediate AED may NOT improve QOL
|
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
