Childhood epilepsy

Dr. O.C. Snead

April 22, 2005.



Diagnostic evaluation:

-         Hx, P/E

-         EEG, MR/CTm, LP, metabolic

-         LP for R/O infection and metabolic causes


Treatment differences:

-         ethosuxamide for absence

-         vigabatrin, ACTH for infantile spasms – no evidence for topiramate

-         partial sz meds same as adults

-         ketogenic diet use, e.g., in LGS (synergistic to VPA for hepatotoxicity)


Side effect differences:

-         learning, behaviour, ADHD side effects

-         higher with PB, possibly levitiracetam

-         idiosyncratic toxic response to lamotrigine is higher

-         idiosyncratic toxic response to VPA is higher

-         idiosyncratic toxic response to felbamate is lower

-         half-life of AEDs is shorter, e.g., TPM, gabapentin


Surgical tx in children:

-         corpus callosotomy

-         hemispherectomy

-         focal cortical excision – temporal

-         focal cortical excision – extratemporal

-         multiple subpial transactions (MSTs)

-         vagal nerve stimulation


First epilepsy surgery is 1881 by Victor Horsley, removal of glial scar from horse accident.


Penfield and Jasper mapped out function with depth electrodes in conscious patients.

But to be used in children, localization had to be determined under anaesthesia. Use of the SSEPs for mapping the primary sensory cortex. Especially important due to shifting functions in children with longstanding epilepsy – cannot rely on anatomical landmarks. From that, development of indwelling electrodes to localize epileptiform activity – first with epidural electrodes, then subdural.


Indications for surgery:

-         Refractory to medical management

-         Localization related or drop attacks

-         Intolerable side effects of AEDs

-         Cognitive deterioration


Medical refractory epilepsy: acceptable seizure control cannot be achieved despite adequate trials of potentially effective drugs at doses associated with acceptable side effects. This is hard to determine.


After failing 2 drugs, chance of responding to 3rd is <5%.

After failing 3 drugs, chance of responding to 4th is <1%.


Goals of surgery:

-         seizure-free state

-         absence of neurological deficit


For success:

-         lateralizeation and localization of seizures

-         localization of the eloquent cortex


Resistance for surgery:

-         natural hx may lead to spontaneous remission – morbidity of seizure effects

-         poor understanding of surgical morbidity and outcomes


Contraindications to surgery:

-         neurodegenerative disease

-         bilateral and/or multifocal regions of onset

-         poor supportive environment

-         significant medical disease

-         non-epileptic events


Corpus callosotomy:

-         complete vs. anterior 2/3

-         works best for drop attacks, usually refractory to AEDs

-         ?risk of post-op mutism in complete callosotomy

-         only fixes the drop attacks, not other seizure types



-         functional hemispherotomy – temporal lobectomy, disconnect the hemisphere, leaving blood supplies (lower morbidity, higher failure rate than complete hemispherectomy)

-         other side needs to be radiologically normal

-         indications are few:

1.      hemimegalencephaly

2.      Rasmussen’s encephalitis

3.      Sturge-Weber Syndrome


Pathway of investigations

  1. hx P/E, video EEG, CT/MRI
  2. decision re candidacy
  3. WADA, fMRI, PET, neuropsych testing
  4. MEG/MSI interictal
  5. design of subdural electrode grid and surgical implantation of grid
  6. recording of seizures and functional mapping by stimulation of electrodes
  7. creation of brain map and removal of grid with focal cortical excision + MSTs


MEG better resolution for localized spikes interictally compared to EEGs. Resolution to a few millimeters. Good correlation with structural and PET abnormalities. Also good concordance with invasive intracranial monitoring (90%). 100% concordance for mapping out sensory cortex compared to invasive monitoring.


Engel classification of postop effects:

I – complete resolution

II – good resolution

III – some help

IV – no help


Seizure control doesn’t reverse cognitive and psychosocial difficulties statistically. May be due to prolonged epilepsy prior to surgery. Also no improvement of quality of life. Strongest predictor for individual change was baseline function. Perhaps only way to get around this is earlier treatment. At least surgery improves future decline.


Multiple subpial transactions:

-         cut tangential tracts to inhibit spread of seizures without affecting deep connections

-         only works in 2ndary zones of epileptogenesis

-         not very effective


Vagal nerve stimulation:

-         an electrode on the vagal nerve stimulated 30 sec q5min

-         mechanism unknown

-         can break clusters of seizures with magnetic stimulation of the VNS


Quality of life issues:

-         school, learning, behaviour, peers, parental understandingm, caregivers, working mothers

-         pregnancy and contraception, catamenial epilepsy, problems with menses, psychosocial adjustment problems

-         schooling, occupation, leisure time activites, housing, social isolation, driving


Body mapping:

-         if the coloured body part could speak, what would it say?

-         Helps to understand child’s experience