Special Populations & SUDEP
Special Populations & SUDEP
What Do You Need to Know?
- VPA is the highest-risk ASM in pregnancy: MCM 10.3%, IQ ↓7–10 points (NEAD), ↑ autism — 2024 AAN mandate: MUST avoid in women of childbearing potential
- Safest ASMs in pregnancy: LTG (2.3–2.9%), LEV (2.4–2.8%), OXC (≤3%); LTG clearance ↑50–100% — monthly levels, target ≥65% preconception
- Elderly = HIGHEST epilepsy incidence (130–140/100,000/yr); stroke most common cause (30–50%); “start low, go slow” with LTG, LEV, or lacosamide
- SUDEP strongest risk factor: ≥3 GTCS/yr (OR 15–23); MORTEMUS: GTCS → PGES → apnea → asystole (respiratory BEFORE cardiac)
- SUDEP prevention: seizure control most important; surgery ↓ risk 80–90%; nocturnal supervision ↓ risk 50–70%
- AAN 2017 (Level B): clinicians SHOULD inform patients about SUDEP; <10% currently receive counseling
Women & Epilepsy: Pregnancy
Teratogenicity Ranking (EURAP / NAAPR Data)
| ASM | MCM Rate | Key Malformations | Dose-Dependent |
|---|---|---|---|
| Valproate | 10.3% | NTDs (1–2%), cardiac, hypospadias, cleft palate | Strong (>600–700 mg/d) |
| Phenobarbital | 6.5% | Cardiac defects, cleft lip/palate | Yes |
| Phenytoin | 6.4% | Cardiac, cleft lip/palate, fetal hydantoin syndrome | Moderate |
| Carbamazepine | 5.5% | NTDs (0.5–1%), cardiac | Yes (>400 mg/d) |
| Topiramate | ~3.9% | Oral clefts (OR 5.4), SGA (18.5%) | Moderate |
| Oxcarbazepine | ≤3% | No specific pattern | Limited data |
| Lamotrigine | 2.3–2.9% | No specific pattern | Weak (>300 mg/d) |
| Levetiracetam | 2.4–2.8% | No specific pattern | Not demonstrated |
2024 AAN Valproate Mandate
- MUST avoid VPA in PWECP to minimize MCMs, NTDs, SGA, and neurodevelopmental harm (autism, ↓ IQ)
- If VPA is the only option → lowest dose (<600 mg/d) + high-dose folic acid (4–5 mg/d)
VPA Neurodevelopmental Effects (NEAD Study)
- IQ ↓7–10 points at ages 3, 4.5, 6 — dose-dependent
- ↑ Autism risk; impaired learning, memory, adaptive functioning
- Folic acid partially mitigates cognitive effects
Pharmacokinetic Changes in Pregnancy
| ASM | Clearance Change | Mechanism | Monitoring |
|---|---|---|---|
| LTG | ↑ 50–100% | Estrogen-induced glucuronidation | Monthly levels; target ≥65% preconception |
| LEV | ↑ 40–60% | Increased renal clearance | Each trimester |
| OXC | ↑ 30–50% | Glucuronidation + renal | Each trimester |
| CBZ / VPA | Minimal change | Autoinduction (CBZ); ↓ protein binding (VPA) | Free VPA levels preferred |
- Postpartum: taper dose increases over 2–3 weeks — LTG toxicity risk within days of delivery
- Seizure freedom ≥9 months preconception → 84–92% chance remaining seizure-free
Folic Acid & Vitamin K
- All PWECP: folic acid ≥0.4 mg/d preconception through pregnancy
- VPA / CBZ: 4–5 mg/d (folate-interfering ASMs)
- Vitamin K 10 mg/d last month if on enzyme inducers (CBZ, PHT, phenobarbital)
Key Registries
| Registry | Key Contribution |
|---|---|
| EURAP | Comparative MCM rates; dose-dependence |
| NAAPR | Specific MCM rates; TPM cleft signal |
| MONEAD | Seizure frequency; PK changes; breastfeeding |
| NEAD | VPA cognitive effects; IQ reduction; folic acid benefit |
Contraception
ASM–Contraceptive Interactions
| ASM Category | Specific ASMs | Effect | Best Contraceptive |
|---|---|---|---|
| Strong inducers | CBZ, PHT, PB, primidone | ↓↓ Estrogen & progestin | LNG-IUD or copper IUD |
| Moderate inducers | TPM ≥200 mg, OXC ≥1500 mg, ESL, clobazam | ↓ Estrogen & progestin | LNG-IUD; avoid OCPs |
| Non-inducing | LTG, LEV, LCM, VPA, GBP, PGB, ZNS, BRV | No effect | Any method |
- LNG-IUD (Mirena): best option on enzyme inducers — local action, not affected by hepatic induction
- Estrogen-containing OCPs ↓ LTG levels 40–60% (active pill weeks; rebound on placebo week)
- Emergency contraception: copper IUD most reliable regardless of ASM
Catamenial Epilepsy
Three Catamenial Patterns (~40% of Women with Epilepsy)
| Pattern | Days | Mechanism | Treatment |
|---|---|---|---|
| C1 — Perimenstrual | Day −3 to +3 | Progesterone withdrawal (most common) | Cyclic progesterone; clobazam |
| C2 — Periovulatory | Day 10–13 | Midcycle estrogen surge | Clobazam or acetazolamide |
| C3 — Inadequate luteal | Day 10 to day 3 next cycle | Low progesterone (anovulatory) | Cyclic progesterone 200 mg TID days 14–28 |
- NIH Progesterone Trial: cyclic progesterone effective for C1 pattern (≥3-fold perimenstrual increase)
- Clobazam 10–20 mg/d for 3–5 days perimenstrually — most practical adjunct
Breastfeeding
ASM Safety During Breastfeeding
- Breastfeeding encouraged on most ASMs — milk transfer << placental transfer
| ASM | Transfer | Safety | Monitoring |
|---|---|---|---|
| LEV, VPA, CBZ | Low | Compatible | Routine observation |
| LTG | Moderate (M:P 0.6–1.0) | Compatible | Monitor rash, sedation |
| Phenobarbital, primidone | Extensive | Caution | Monitor sedation, poor feeding |
| Zonisamide | Extensive (M:P 0.9–1.0) | Caution | Monitor sedation |
Elderly Epilepsy
Epidemiology & Etiology
- HIGHEST incidence: 130–140/100,000/yr (exceeds childhood peak)
- Stroke = #1 identifiable cause (30–50%); ~50% cryptogenic
- Other: neurodegeneration (10–20%), tumors (5–10%), TBI, metabolic
- NCSE in 28% of elderly unexplained delirium on cEEG
- Transient epileptic amnesia: may be earliest manifestation of Alzheimer disease
- Wicket spikes and BETS: benign variants commonly over-interpreted in elderly
ASM Selection: “Start Low, Go Slow”
| Preferred | Rationale |
|---|---|
| Lamotrigine | Best tolerability (VA Cooperative Study); minimal cognitive effects |
| Levetiracetam | Most used; rapid titration; IV; no interactions; renal dosing |
| Lacosamide | Emerging; IV; minimal cognitive effects; check ECG (PR interval) |
AVOID in the Elderly
- PHT: nonlinear kinetics, interactions, osteoporosis, cognitive impairment
- CBZ: CYP3A4 inducer, hyponatremia, interactions
- Phenobarbital: sedation, falls, enzyme induction
- TPM: cognitive impairment — worst in elderly with baseline vulnerability
- Seizure freedom in 60–70% with monotherapy at low–moderate doses
Drug Interactions in the Elderly
| Medication | Interaction | Consequence |
|---|---|---|
| Warfarin | PHT, CBZ, PB ↑ metabolism | Subtherapeutic INR → stroke |
| DOACs | CBZ, PHT: CYP3A4/P-gp induction | ↓ DOAC levels → thromboembolism |
| Statins | CYP3A4 induction | ↓ Statin levels |
| Donepezil | Minimal direct | May modestly ↓ seizure threshold |
Epilepsy–Dementia Bidirectional Link
- AD: 6–10× seizure risk; amyloid-beta ↑ neuronal excitability
- New-onset epilepsy may predict future dementia (ARIC study)
- Transient epileptic amnesia may be earliest AD manifestation
- Subclinical hippocampal seizures in AD (Lam et al., 2017)
- SE in elderly: mortality 20–40%; prefer LEV IV over PHT
SUDEP
Overview
- Leading cause of epilepsy-related premature mortality; 7.5–17% of all epilepsy deaths
- Peak age 20–45 years; 55–70% occur during sleep; prone position in majority
Incidence by Population
| Population | Incidence (/1,000 pt-yr) |
|---|---|
| Children | 0.22 |
| Adults (community) | 1.0–1.2 |
| Drug-resistant | 3.0–6.0 |
| Dravet syndrome | Up to 15 |
Risk Factors
| Risk Factor | Odds Ratio | Notes |
|---|---|---|
| ≥3 GTCS/yr | OR 15–23 | STRONGEST |
| Nocturnal seizures | OR 2.6–5.0 | 55–70% SUDEP during sleep |
| Prone position | OR 4.0–6.0 | 11/16 MORTEMUS cases prone |
| No supervision | OR 3–5 | Modifiable; supervision = OR 0.4 |
| Polytherapy (≥3 ASMs) | OR 2.5–4.0 | Proxy for severity |
| Male sex | OR 1.3–1.8 | — |
| Duration ≥15 yr | OR 1.9 | — |
| Intellectual disability | OR 2.0–3.0 | — |
| Subtherapeutic ASM levels | OR 2.0–3.0 | Found in 30–50% SUDEP autopsies |
MORTEMUS Study — Mechanism
- 16 SUDEP + 9 near-SUDEP on video-EEG in EMUs; ALL 16 followed a GTCS
- Sequence: GTCS → tachypnea → PGES → apnea → bradycardia → asystole
- RESPIRATORY BEFORE CARDIAC in all cases
- 11/16 found prone; PGES >50 s in all cases
- All 9 near-SUDEP survived with immediate intervention
SUDEP Prevention & AAN Guideline
Prevention Strategies
| Strategy | Evidence | Details |
|---|---|---|
| Seizure control | Most important | Eliminate GTCS; optimize ASMs; surgery if drug-resistant |
| Epilepsy surgery | ↓ Risk 80–90% | Post-surgical seizure-free: 0.4–0.7/1,000 pt-yr |
| Nocturnal supervision | OR 0.4 (50–70% ↓) | Bedroom sharing; capable observer |
| Detection devices | Emerging | Empatica (wrist, FDA); Nightwatch (arm, CE); mattress sensors |
| Avoid prone sleeping | Modifiable | Anti-suffocation pillows as adjunct |
| ASM adherence | Subtherapeutic in 30–50% autopsies | Pill organizers; reminders; long-acting formulations |
AAN SUDEP Guideline (2017)
- Level B: clinicians SHOULD inform patients and caregivers about SUDEP, especially with frequent GTCS
- Level B: GTCS frequency is strongest risk factor; seizure freedom = lowest risk
- Level C: may advise nocturnal supervision
- <10% of patients report receiving SUDEP counseling
- When: at diagnosis, when changing treatment, when control worsens, periodically
Board Pearls
💎 Board Pearl
- VPA teratogenicity hierarchy: VPA 10.3% > PB 6.5% > PHT 6.4% > CBZ 5.5% > TPM 3.9% > OXC ≤3% > LTG 2.3–2.9% ≈ LEV 2.4–2.8%. Know this ranking cold.
- LTG clearance ↑50–100% in pregnancy via estrogen-driven glucuronidation — monthly levels; target ≥65% preconception. Most tested PK change in pregnancy.
- OCPs ↓ LTG levels 40–60%. Bidirectional: enzyme inducers ↓ OCP efficacy, AND estrogen ↓ LTG levels. LNG-IUD avoids both problems.
- MORTEMUS sequence: GTCS → PGES → apnea → asystole. RESPIRATORY before CARDIAC. Highest-yield SUDEP fact.
- ≥3 GTCS/yr = strongest SUDEP risk factor (OR 15–23). Epilepsy surgery ↓ SUDEP 80–90%.
- Elderly = highest epilepsy incidence (130–140/100,000/yr). Stroke #1 cause. Avoid PHT/CBZ. Use LTG (best evidence), LEV (fastest), lacosamide.
- Wicket spikes and BETS: benign EEG variants over-interpreted in elderly. Do NOT start ASMs based on these alone.
Clinical Pearls
Clinical Pearl
Postpartum LTG toxicity is a critical safety issue. Clearance returns to baseline within days of delivery. If pregnancy dose increases are not tapered over 2–3 weeks, supratherapeutic levels cause diplopia, ataxia, and nausea. Board-relevant scenario.
Clinical Pearl
NCSE should be on the differential for any elderly patient with unexplained delirium. cEEG detected NCSE in 28% of elderly with unexplained altered mental status. Low threshold for cEEG when confusion is disproportionate to medical causes.
Clinical Pearl
New-onset epilepsy in the elderly may be the earliest sign of Alzheimer disease. ARIC study: late-onset epilepsy of unknown etiology predicts subsequent dementia. Baseline MoCA at diagnosis; reassess annually. Choose LTG, LEV, or lacosamide.
References
- Pack AM, et al. Teratogenesis, perinatal, and neurodevelopmental outcomes after in utero ASM exposure (2024 AAN/AES/SMFM guideline). Neurology. 2024;102(11):e209279.
- Tomson T, et al. Comparative MCM risk with 8 ASMs (EURAP). Lancet Neurol. 2018;17(6):530–538.
- Meador KJ, et al. Fetal ASM exposure and cognitive outcomes at age 6 (NEAD). Lancet Neurol. 2013;12(3):244–252.
- Pennell PB, et al. Seizure frequency and ASM therapy during pregnancy (MONEAD). N Engl J Med. 2020;383(26):2547–2556.
- Pennell PB, et al. ASM concentrations during pregnancy (MONEAD). JAMA Neurol. 2022;79(4):370–379.
- Herzog AG. Catamenial epilepsy: NIH Progesterone Trial update. Seizure. 2015;28:18–25.
- Ryvlin P, et al. Cardiorespiratory arrests in EMUs (MORTEMUS). Lancet Neurol. 2013;12(10):966–977.
- Harden C, et al. SUDEP incidence rates and risk factors (AAN guideline). Neurology. 2017;88(17):1674–1680.
- Devinsky O, et al. SUDEP: epidemiology, mechanisms, prevention. Lancet Neurol. 2016;15(10):1075–1088.
- Langan Y, et al. Case-control study of SUDEP. Neurology. 2005;64(7):1131–1133.
- Johnson EL, et al. Dementia in late-onset epilepsy (ARIC). Neurology. 2020;95(24):e3248–e3256.
- Rowan AJ, et al. Geriatric epilepsy: GBP vs. LTG vs. CBZ (VA Cooperative Study). Neurology. 2005;64(11):1868–1873.
- Tomson T, et al. Management of epilepsy in pregnancy (ILAE Task Force). Epileptic Disord. 2019;21(6):497–517.
- Bensalem-Owen MK. Reproductive health in epilepsy. Continuum. 2025;31(1):214–231.
- Beniczky S, et al. Automated seizure detection wearables (ILAE/IFCN guideline). Clin Neurophysiol. 2021;132(5):1173–1184.