Epilepsy Surgery Indications

What You'll Learn

Drug-resistant epilepsy = failure of 2 appropriate ASMs at adequate doses (ILAE definition); refer early — do NOT wait for 5+ ASM failures
Best surgical candidates: focal onset seizures with an identifiable MRI lesion and concordant presurgical data
ERSET trial: early surgery for mTLE = 73% seizure-free vs. 0% with continued medical therapy — surgery is superior, not a last resort
Temporal lobectomy is the most common and most successful epilepsy surgery; 60–80% Engel I for MTLE with hippocampal sclerosis
Presurgical workup: Phase I (video-EEG, MRI epilepsy protocol, PET, neuropsych) → Phase II (SEEG or grids) if non-concordant
Palliative options: corpus callosotomy for drop attacks (LGS); hemispherotomy for hemispheric epilepsy syndromes (Rasmussen, Sturge-Weber)
Engel classification: Class I = seizure-free (Ia = completely free, Ib = auras only); the standard outcome measure for epilepsy surgery

HighYield Pearls

ILAE drug-resistant epilepsy definition: failure of 2 well-chosen, well-tolerated ASMs — NOT 3, NOT 5; refer for surgical evaluation immediately, do not wait years
mTLE with hippocampal sclerosis → standard ATL (or LITT) → 60–70% Engel I — the prototype “ideal” surgical candidate
FCD type IIb with transmantle sign + balloon cells → best surgical candidate among non-HS pathologies (~70–80% Engel I); complete resection is the critical predictor
Non-lesional focal epilepsy → needs SEEG to localize seizure onset zone; lower Engel I rate (~30–40%)
Rasmussen encephalitis, Sturge-Weber with hemiplegia, hemimegalencephaly, large MCA stroke → functional hemispherotomy; younger patients have greater contralateral plasticity
Disabling drop attacks in LGS / Dravet / FIRES → corpus callosotomy (palliative; ~70–80% drop reduction); NOT a cure
Bilateral mesial temporal or eloquent-cortex onset → RNS (closed-loop, ≤2 leads, ~75% median reduction at 9 yr)
Pre-op memory lateralization → Wada or fMRI required before dominant temporal resection to risk-stratify verbal memory decline
Post-op ASM management: continue initially; consider taper at 2 yr seizure-free — counsel patient on recurrence risk with withdrawal
Multidisciplinary epilepsy center is mandatory: neurologist + neurosurgeon + neuropsych + neuroradiology + EEG — surgical decisions are not made in isolation

🔍 Quick ReferenceCandidate selection · Lesional / pathology · Outcomes

Candidate selection

Failure of 2 well-chosen, well-tolerated ASMs → ILAE drug-resistant epilepsy — refer for surgical evaluation
Non-lesional focal epilepsy, MRI-negative, or eloquent-cortex onset → SEEG (or subdural grids) for invasive localization
Dominant temporal resection planned → Wada or fMRI for memory/language lateralization
Tuberous sclerosis with multifocal tubers → SEEG to identify dominant epileptogenic focus
Cryptogenic focal epilepsy, MRI-negative → MEG + functional brain mapping to localize

Lesional / pathology

Hippocampal sclerosis (MTLE-HS) → ATL or LITT — 60–70% Engel I
FCD type IIb with transmantle sign + balloon cells → lesionectomy — best non-HS surgical candidate (~70–80% Engel I)
Cavernoma → resection of lesion + perilesional gliosis / hemosiderin ring
DNET, ganglioglioma, low-grade astrocytoma → lesionectomy — 75–90% Engel I (highest of any pathology)
Encephalomalacia post-stroke / post-traumatic → resection of gliotic scar + epileptogenic zone
Rasmussen encephalitis / Sturge-Weber / hemimegalencephaly / large perinatal MCA stroke → functional hemispherotomy

Outcomes / pearls

MTLE-HS → 60–70% Engel I; FCD IIb with concordant data → ~70–80% Engel I
Low-grade tumor (DNET / ganglioglioma) → 75–90% Engel I — the best outcomes
Non-lesional / MRI-negative → 30–40% Engel I — the worst outcomes
Corpus callosotomy for atonic drops in LGS → 70–80% drop-attack reduction (palliative, not curative)
RNS at 9 yr (Nair 2020) → ~75% median seizure reduction — accumulating effect over time
VNS → “50/50 rule” (~50% responder rate, ~50% seizure reduction); observational SUDEP risk reduction
ANT-DBS (SANTE) → ~75% median seizure reduction at 7 yr; watch for depression + memory complaints
Engel classification (I seizure-free, II rare, III worthwhile, IV no benefit) → ILAE 2012 classification is the alternative outcome scale

Surgical Candidacy — When to Refer

Drug-Resistant Epilepsy (ILAE Definition)

Definition: failure to achieve sustained seizure freedom after adequate trials of 2 tolerated, appropriately chosen ASMs (monotherapy or combination)
After 2 ASM failures, probability of seizure freedom with each additional agent drops to ~5%
Refer to a comprehensive epilepsy center as soon as drug resistance is established
Average delay from drug resistance to surgery referral is 10–20 years — this is unacceptable given Level 1 evidence

Landmark Trials Supporting Early Surgery

Trial	Year	Design	Key Result
Wiebe et al.	2001 (NEJM)	RCT: ATL vs. medical therapy for TLE	58% vs. 8% seizure-free at 1 year; NNT = 2
ERSET	2012 (JAMA)	RCT: early surgery vs. continued medical therapy	73% vs. 0% seizure-free at 2 years

NNT of 2 (Wiebe; ARR 50%) is among the largest treatment effects in clinical neurology
ERSET caveat: stopped early for slow accrual (n=38 of planned 200); results are hypothesis-generating despite the striking effect size
Earlier surgery → better cognitive outcomes, better psychosocial functioning, lower SUDEP risk

Ideal Candidates

Focal onset seizures with an identifiable epileptogenic focus
MRI-visible lesion (hippocampal sclerosis, FCD, low-grade tumor, cavernoma)
Concordance across all presurgical modalities (EEG, MRI, PET, semiology, neuropsych)
mTLE with hippocampal sclerosis = the “ideal” surgical candidate

Relative & Absolute Contraindications

Contraindication	Type	Notes
Primary generalized epilepsy	Absolute (for resective surgery)	May still be considered for palliative procedures (callosotomy, VNS)
Progressive/degenerative cause	Absolute	Neurodegenerative or metabolic etiologies with diffuse involvement
Bilateral independent foci	Relative	May proceed if one focus is clearly dominant or if palliative surgery is planned
Focus in eloquent cortex	Relative	Cortical mapping (SEEG, fMRI) needed; consider neuromodulation (RNS) as alternative
Severe psychiatric comorbidity	Relative	Active psychosis or suicidality may need stabilization first; not absolute

Board Pearls

Drug-resistant epilepsy = failure of 2 ASMs (not 3, not 5) — boards test this ILAE threshold repeatedly
After 2 ASM failures, each additional agent adds only ~5% chance of seizure freedom — surgery should be discussed, not deferred
Primary generalized epilepsy is NOT a candidate for resective surgery but may benefit from callosotomy or neuromodulation

Presurgical Evaluation — The Neurology Perspective

Phase I (Noninvasive)

Modality	What It Shows	Key Points
Video-EEG monitoring	Seizure-onset zone, semiology, interictal discharges	Gold standard; capture ≥3–5 habitual seizures; typically 5–14 day admission
3T MRI (epilepsy protocol)	Structural lesion (HS, FCD, tumors, cavernomas)	NOT a “routine brain MRI”; includes 3D T1/FLAIR, coronal T2 perpendicular to hippocampus, SWI
FDG-PET	Interictal hypometabolism at epileptogenic zone	80–90% sensitivity for mTLE; more sensitive than MRI for subtle lesions
Ictal SPECT (SISCOM)	Ictal hyperperfusion at seizure-onset zone	Ideally inject within 20–30 sec of clinical onset; injection >45–60 sec risks capturing propagation patterns rather than onset
MEG / MSI	Magnetic source imaging of interictal discharges	Most useful in MRI-negative cases; detects sulcal cortex better than EEG
Neuropsychological testing	Baseline cognition; lateralization of language/memory	Verbal memory deficit → left temporal; visuospatial → right temporal

Phase II (Invasive Monitoring)

Indications: non-concordant Phase I data, MRI-negative cases, seizure onset near eloquent cortex, bilateral independent onsets
Required in 30–40% of surgical candidates

SEEG vs. Subdural Grids

Feature	SEEG (Stereo-EEG)	Subdural Grids
Implantation	Stereotactic via twist-drill holes; robot-assisted	Open craniotomy
Spatial coverage	Deep structures (hippocampus, insula, cingulate); bilateral feasible	Cortical surface; limited deep access
Complications	Hemorrhage 1–4%; infection 1–2%	Overall 10–15% (hemorrhage, infection, edema)
Cortical mapping	Limited by electrode geometry	Excellent for motor/language mapping
Current trend	Has largely replaced grids in North America	Declining; reserved for specific cortical mapping needs

Language & Memory Lateralization

fMRI: has replaced the Wada test for language lateralization (>90% concordance); noninvasive, widely available
Wada test (intracarotid amobarbital): still needed for memory lateralization, especially before dominant temporal resection
Wada activation pharmacology: sodium amobarbital (Amytal) is no longer manufactured; methohexital and etomidate (eSAM — Etomidate Speech and Memory Test) are the modern substitutes at most centers. Etomidate also activates interictal epileptiform discharges intraoperatively.
Left hemisphere = language dominant in 95% of right-handers, ~70% of left-handers

The Concordance Principle

All modalities must point to the same focus — this is the fundamental principle of presurgical evaluation
Full concordance (semiology + EEG + MRI + PET + neuropsych) → >70% seizure-free outcome
Discordance → need Phase II investigation or may not be a surgical candidate
Final decision made at multidisciplinary epilepsy surgery conference

Board Pearls

FDG-PET = interictal HYPOmetabolism; Ictal SPECT = ictal HYPERperfusion — opposite findings, both localizing the epileptogenic zone
fMRI replaces Wada for language; Wada still needed for memory lateralization (especially left TLE)
SEEG has replaced subdural grids at most centers: lower complications, better deep structure access, bilateral sampling

🔒

Continue reading — sign in

The full note has more clinical pearls, tables, and board-focused tips. Free account, no fee.