This document discusses non-pharmacological treatment options for epilepsy, including epilepsy surgery, dietary treatments, and neurostimulation. It provides information on when these alternatives should be considered, the options available, and for which patients they may be suitable. Key points include: epilepsy surgery may be underused in the UK and can provide seizure freedom or reduction; the ketogenic diet has been shown to reduce seizures in about 15-30% of patients after 6 months depending on the study; and corpus callosotomy and vagus nerve stimulation are established palliative procedures for treatment-resistant generalized seizures.

1. Epilepsy Treatment –
alternatives to anti-epileptic drugs
Marios Kaliakatsos
Consultant Paediatric Neurologist
Great Ormond Street Hospital for Children, London

2. Outline
• When should non-pharmacological treatment options be considered ?
• Which options are available ? & For which patients ?
• Epilepsy Surgery
• Dietary Treatment of Epilepsies
• Neurostimulation (VNS, DBS)
• Alternative medication treatments (cannabinoids)
• The future

5. Drug resistant epilepsy is defined as failure of
adequate trials of two tolerated, appropriately
chosen and used antiepileptic drug schedules
(whether as monotherapies or in combination)
to achieve sustained seizure freedom
ILAE 2010

6. How common is it ?
• Epilepsy clinic setting:
• 25% not seizure free for at least 1year
• (n=1,098, age 9-23y), Brodie et al, Neurology 2012
• Population based setting:
• Adults (>15 y): 16% (Picolet et al, 2008)
• Children:
• Netherlands: DSEC: 9 % (no > 3 m remission phase + failure ≥ 2 AEDs (Geerts et al 2010)
• USA, Connecticut cohort 23% failed at least 2 AEDs (Berg AT et al, 2015)
• Type of epilepsy and aetiology:
• Temporal lobe epilepsy:
• 37 % (Dlugos et al, 2001)
• 69% (Spooner et al 2006)

7. • “Individuals with poorly controlled epilepsy may benefit from referral to
a tertiary centre and further assessment, which may include assessment
for epilepsy surgery.”
• “…epilepsy surgery may be underused as a treatment modality for
poorly controlled epilepsy in the UK owing to suitable individuals not
being referred to a tertiary centre.”
The Epilepsies , The diagnosis and management of the epilepsies in adults and children in primary
and secondary care , January 2012
NICE clinical guideline 137
www.nice.org.uk/cg137

8. Aims of epilepsy surgery
• Primary outcome
• Seizure freedom/reduction
• Secondary outcome aims
• Optimising potential for neurodevelopmental gains
• Behavioural improvement (? downstream effect – reduction of AED
burden)
Slide courtesy of Helen Cross

9. Presurgical Evaluation
Identification of epileptogenic zone
→generation of focal seizures
→removal necessary for seizure freedom
Lesion

10. Presurgical Evaluation
Eloquent cortex
(cortical stimulation, fMRI)
Functional deficit
zone
(Neuropsychology,
FDG-PET)
Lesion (MRI)
Ictal onset zone
(video telemetry, SPECT)
Irritative zone
(interictal EEG)
Slide Courtesy of Ronit Pressler

11. Multi-modality pre-surgical work up
• Core investigations:
• Scalp video EEG (video-telemetry)
• Optimised MRI
• Neuropsychology + Neuropsychiatry
• Extended work up step1:
• PET, SPECT
• FMRI
• language, motor (eloquent cortex)
• spike activated (irriative zone)
• MEG
• Specialised MRI (DTI)
• Extended work up step 2
• Invasive EEG recording
• Subdural grids, depth electrodes
• Stereo tactic EEG (stereotactic depth electrode implantation)

12. Ictal SPECT
Single photon emission tomography
• Radio tracer injection at onset of
seizure
• Visualisation of cerebral blood
flow
• Seizure activity - ↑cerebral
metabolism coupled to
↑hyperperfusion
• SISCOM – substraction – ictal
from interictal data and co-
registration with MRI
Ictal Interictal

13. FDG-PET –
2-deoxy 18F fluoro-D-glucose - Positron emission tomography
• Visualisation of interictal
hypometabolism
• Mechanism unknown
• Possible reversible state associated
with interictal inhibitory processes

14. FDG-PET co-registered with MRI
Enhances detection of lesions
especially - FCD (FCD2)
(Salamon, Kung et al. 2008)

15. 6
4
8
1
5
7
I
F
S
SF
S
T
S
G&
R
D
P
D
A
Superior
Inferior
AnteriorPosterior
images courtesy of Kelly St Pier
Invasive EEG recording
Subdural grid implantation
• to delineate seizure onset zone:
• MR lesion negative – other investigations suggesting focal onset
• MR lesion positive cases:
• with unclear demarcation of borders or/and some discordance
of other investigations (i.e. scalp video EEG)
• Some cases of tuberous sclerosis
• Mapping of eloquent cortex extra-operative required

16. Invasive EEG Monitoring: SEEG
Stereotactic depth electrode implantation
• Better tolerated
• Exploration of wider cortical areas possible
• Deeper cortical areas can be explored
• Detailed planning, based on hypothesis
• Thermocoagulation of ictally most involved
contacts possible (‘proof of principle’)
• Limitations – functional cortical mapping – i.e.
language

17. Large hemispheric Lesions
Acquired:
MCA territory infarct (here left)
Developmental cortical malformation:
Hemimegalencephaly (here right)
Surgical procedure considered
Functional
hemispherectomy
→Subtotal anatomical
resection with full
disconnection of the
abnormal hemisphere

18. Focal lesions
Mesial temporal sclerosis
(here right)
Tumor
(here low grade
glioneuronal tumor)
Focal cortical dyslplasia
(here left superior + middle frontal gyrus)
Tuberous sclerosis
Surgical procedures considered:
• Lesionectomy
• Lobectomy
• Disconnection (i.e. temporal-
occipital-parietal
disconnection)

19. Percent
Pediatric Epilepsy Surgery Patients < 18 Years
(ILAE Survey 2004; 20 Centers Europe, Australia, & USA; n=413)
42.4%
Harvey et al., Epilepsia 2008;49:146-155
0 10 20 30 40 50
Vascular
Rasmussen Syndrome
Sturge-Weber
Hypo. Hamartoma
Tuberous Sclerosis
Gliosis/Normal
Hippocampal Sclerosis
Atrophy/Stroke
Tumor
Cortical Dysplasia
Slide courtesy of Helen Cross

20. Epilepsy surgery
? Worth exploring if MRI lesion negative
Téllez-Zenteno et al 2010, Epilepsy Res 89, 310-318
systematic literature review: 40 papers, 697 patients non-lesional, 2860
patients lesional >= 1 year follow up
Paediatric subgroup:
Meta analysis of temporal epilepsy surgery
20 studies – adults and children – 1657 patients
Lesion present vs no lesion (MRI or pathology)
 Odds of seizure freedom 2.7 higher
(95% CI 2.1-3.5, p < 0.001)

21. Corpus – Callosotomy (CC)
• Palliative procedure for pharmaco-
resistant generalised seizures
• Most effective for drop attacks
• Often tolerated better in younger
children (≤ 12y)
• Up to 88% worthwhile reduction in
seizures with complete CC
compared to 56: with anterior CC
only (Graham D et al , Epilepsia 2016)
• Transient disconnection syndrome*
more frequent with complete CC
(*decreased use, dyspraxia of non-dominant arm/hand,
decreased spontaneous speech , lethargy, urinary
incontinence)
Before CC After CC

22. Corpus Callosotomy
GOSH and Children’s Hospital, Westmead, Sydney
Graham D et al Epilepsia, 57(7):1053-1069
• 55 children under 18 years
(mean age @ surgery 10.4 y)
• LGS, West and Ohtahara
syndrome most commonly
• No pre-surgical predictors of
outcome
• 47% had rare or no drop attacks
(median follow-up 36 months)
• If drop attacks do return, most
likely in first 12 months
• Fewer injuries after surgery
• Reduced drug-burden

23. ©2011 American Academy of Neurology. Published by Lippincott Williams & Wilkins, Inc. 2
Long-term intellectual outcome after temporal lobe surgery
in childhood
Skirrow, C; Cross, JH; MD, PhD; Cormack, F; Harkness, W; Vargha-Khadem, F; Baldeweg, T
Neurology. 76(15):1330-1337, April 12, 2011.
DOI: 10.1212/WNL.0b013e31821527f0
Figure 3 Longitudinal profile of IQ changesPreoperative
to postoperative full-scale IQ changes across time after
surgery, shown for subsequent 2-year periods (positive
values denote IQ gains). Note that only a proportion of all
patients were assessed during each time period
(indicated as a percentage of the total group).
*Significant changes (one-sample t tests, p <= 0.05).
Predictors of FSIQ
change:
• Cessation of AEDs
• Pre-surgical FSIQ
(greater improvement
at lower end of
spectrum compared to
average and high
average)

24. • N=66
• Neuropsychological
evaluation
• Presurgical:
• 63 (93%), median 12m before
• Post surgical
• 36 (54%), after median of
18.6 m
Presurgical FSIQ only factor predicting postsurgical FSIQ

25. • Factors independently predicting low pre-surgical FSIQ
• Younger age at epilepsy onset (< 3 years)
• Duration of epilepsy
• Aetiology
• Low grade tumors better function than FCD + other pathologies
• Gender (worse out come in females)

26. Conclusions
• Primary aim remains seizure freedom
• Secondary aims:
• improvement in psychosocial functioning
• Improvement of life quality
• Refer early for evaluation
• Duration of evaluation process
• Benefit from shorter duration of epilepsy and reduction of AED burden
• Detailed evaluation required to determine
• Individual outcome goals (benefits)
• risks (i.e. possible impairments – motor, language, visual filed)
Slide with amendments - courtesy of Helen Cross

29. What are ketogenic diets ?
Normal Diet
Energy from fat rather than
carb’s

30. MCT Ketogenic Diet
Modified Atkins Diet (1:1 ratio)
Classical Ketogenic Diet 4:1 ratio (LCT)
Normal Diet
E. Neal, 2012
Dietary Treatment of Epilepsy, Wiley-Blackell

31. Classical Ketogenic Diet (CKD):
(Long Chain Triglycerides)
• Using standard food for
composition of meals plans:
2:1, 3:1 or 4:1 -
fat: (carbohydrate + protein) ratio
• Up to 90% of total calories from
fat
• Meal/snack recipes, all in correct
ratio

32. Ketogenic feeds
Nutritionally complete, provides
calories and protein for growth as well
as vitamins and minerals

33. Medium Chain Triglyceride (MCT)
Ketogenic Diet
• 40-60% of daily calorie intake as
MCT oil / MCT food product
(overall 75% fat)
• less carbohydrate restricted
(15-20% of total calories)
• (Possibly) Greater choice of foods

34. •Supplements
• Vitamin, minerals and
trace elements

35. Modified Ketogenic Diet (MKD)
• Carbohydrate restriction (10-20g/day)
• Encourages consumption of high fat food
• No limit on proteins
• (No limit on total calories)
• Fat : (carbohydrate + fat ) 1:1 ratio
• 70% fat, 25% protein, 5% carbohydrates

36. Low GI diet
• ‘Glycaemic Index’
• Fewer fluctuations in glucose lead to effective sz control
• Carbohydrate restriction – 40-60 gm/day

37. Efficacy in childhood epilepsy
D Keene Ped Neurol 2006;35:1-5
A systematic review
• 26 studies;14 met criteria for inclusion, mostly
retrospective, no control group
• Outcome measures degree of seizure control,
duration patient remained on diet, occurrence of
adverse events
• Total collective population 972 patients
• At 6m
• 15.6% (CI 10.4-20.8) seizure free
• 33.0% (CI 24-41.8) >50% reduction

38. Which KD type works better ?
• Classical KD versus MCT
• Neal et al , Epilepsia 50(5):1109-1117, 2009 (45 on Classical, 49 on
MCT; age 2-16y)
• both KD types have comparable efficacy (no significant difference between
mean percentage of baseline seizures at 3,6 and 12 months)
• Classical KD versus MKD
• Kim et al, Epilepsia, 57(1):51-58, 2016,(51 on Classical, 53 on MKD; age 1-
18y)
• Mean percentage of baseline seizures:
• After 3 months: Classical KD 38.6%, MKD 47.9%
• After 6 month: Classical KD 33.8%, MKD 44.6%
• Difference not statistically significant in overall group

39. KD in specific subgroups
• Epilepsy Syndromes:
• Infantile spasms
(Hong, Turner et al. 2010; Nordli, Kuroda et al. 2001)
• Dravet Syndrome
(Caraballo, Cersosimo et al. 2005; Dressler, Trimmel-Schwahofer et al. 2015)
• Myoclonic – Atonic Epilepsy
(Stenger E et al 2017; Wiemer-Kruel, Haberlandt et al. 2017)
• Lennox-Gastaut Syndrome
(Lemmon M et al 2012; Zhang et al 2016)
• Glut1 deficiency
(Leen et al Brain 2010)

40. Ketogenic Diet - Side Effects
• Gastrointestinal symptoms:
• nausea, vomiting (worsening of Gastro-oesophageal Reflux), constipation
• Low blood Sugar (occasionally in initiation phase)
• Excess ketosis – acidosis (initiation phase)
• Renal stones (3-6%)
• Risk factors: young age, hypercalciuria, (tx with carbonic anhydrase
inhibitors: Topiramate, Zonisamide)
• Prevention – potassium citrate (alkalinisation of urine)
reduction from 6.7 to 0.9 % (McNally et al, Pediatrics, 2009)
• Increased Bruising (Berry-Kravis et al, Ann Neurol 2000)
• Weight loss, Inadequate growth
• Pancreatitis
• Hyperlipidaemia
• Decreased bone density – fractures (Long-term treatment)

41. When to consider KD treatment
• Seizures despite of AED treatment
(usually - failure of ≥ 2 AEDs)
• Poor tolerance to AEDs
• (Rare) Metabolic disorders affecting
• transport of glucose from blood into brain
• Glut 1 transporter deficiency syndrome
• Metabolism of glucose
• Pyruvate dehydrogenase deficiency

42. When would be the KD be contraindicated ?
• Metabolic conditions
• Beta-Fatty oxidation defects
• Familial hyperlipidaemia
• Organic acidurias
• Pyruvate carboxylase deficiency (lactic acidosis)
• Relative contraindications
• Feeding difficulties (food refusal)
• Dysphagia (alternative feeding route: NG tube or PEG)
• Severe gastro-oesophageal reflux (frequent vomiting)

43. Ketogenic Diet – Duration of Treatment
• 3.5 months to assess efficacy
• Consensus statement Kossoff et al Epilepsia, 50(2):304–317, 2009
• First effects after 2 weeks
• Duration of treatment
• Initially 2 years (than taper diet)
• in sz free patients sz control often maintained
20% relapsed after discontinuation (Martinez et al 2007,
Epilepsia)
• In Glut 1 deficiency syndrome:
• Until after puberty (or longer into adulthood ? Transition to
MAD)

44. Mechanisms – hypotheses
Bough &Rho Epilepsia 48 (1):43-58, 2007
Rho & Stafstrom Epilepsy Research 2011
• Anti-epileptic effect not only mediated by ketone bodies – but by
adaptive metabolic processes induced by ketosis
• Effects mediated by polyunsaturated fatty acids
• Ketosis induces shifts in brain amino acid handling favouring GABA
production
• Suppression of seizures mediated by adenosine acting on adenosine
A1 receptors

45. C 10 (decanoic acid)
Potential explanation why MCT diet works
• Increases number and function of mitochondria in cells
• Hughes SD et al, J Neurochem, 2014
• Chang et al, Neuropharmacolgy 2013
• C10 (also C9 ) decrease of epileptiform discharges - in vitro model
• Can suppress epileptiform activity by blocking AMPA receptors
(receptor for excitatory neurotransmitter Glutamate)
• Chang et al, Brain 2016
• On-going first study to evaluate feasibility of new MCT food product
(higher percentage of C10 )
Chief Investigator: Prof M Walker

46. Conclusions
• Ketogenic diet can be effective in proportion of patients & is well
tolerated
• MCT, Classical diet and MKD all beneficial
• Dietary treatments require close monitoring and input from a team
of physician, specialist dietitian and epilepsy nurse
• Appropriate treatment goals should be set prior to intervention
• MCT – with higher proportion of decanoic acids shows promising
results – further RCT required to evaluate clinical efficacy and long-
term effects

47. NICE clinical guideline 137
www.nice.org.uk/cg137*
179. Vagus nerve stimulation is indicated for use as an adjunctive
therapy in reducing the frequency of seizures in children and young
people who are refractory to antiepileptic medication but who are not
suitable for resective surgery. This includes children and young people
whose epileptic disorder is dominated by focal seizures (with or without
secondary generalisation) or generalised seizures. [2004, amended
2012]
*The Epilepsies , The diagnosis and management of the epilepsies in adults and
children in primary and secondary care , January 2012

48. VNS Therapy
• Pulse generator device is
implanted surgically in the left
chest/axilla
• Electrodes are tunneled
beneath the skin from the
pulse generator to the left
vagus nerve in the neck.
• The vagus nerve is stimulated
periodically at the site of the
neck.
• Surgical scars left axilla and
left neck
Slide courtesy of Sophia Varadkar

49. VNS Therapy: the components
LeadPulse generator
Electrodes
Slide courtesy of Sophia Varadkar

50. Programming System Components
Handheld Computer
• Platform for Programming
Software
Programming Wand
• Accessory to programming
handheld computer
• Communication tool between
Programming Software and
Pulse Generator
Slide courtesy of Sophia Varadkar

51. Vagus Nerve Stimulation Therapy
• Non-pharmacological therapy for
epilepsy
• Repeated electrical stimulation of
the left vagus nerve by the pulse
generator device
• 3 areas of benefit for seizures
• Acute abortive
• Acute prophylactic
• Chronic progressive
• Other benefits?
• Quality of life
• Mood, behaviour, alertness,
memory, school
Slide courtesy of Sophia Varadkar

52. Potential mechanisms of action
• Animal models
• Early neurophysiological studies indicated desynchronisation in cats
• Locus coeruleus mediates anti-seizure effect of VNS
• Human studies
• Changes in cerebral blood flow;
• Changes in CSF neurotransmitters
Slide courtesy of Sophia Varadkar

53. VNS Effectiveness Over Time
Morris GL, Mueller WM. Neurology. 1999;53(7):1731-1735.
Slide courtesy of Sophia Varadkar

54. VNS Therapy Seizure-Free Rates
1. Renfroe JB and Wheless JW. Neurology 2002;59(suppl 4):S26-S30. 2. Helmers SL, et al. Neurologist
2003;9:160-4. 3. De Herdt V, et al. Eur J Paediatr Neurol 2007;11:261-9. 4. Amar AP, et al. Neurosurgery
2004;55:1086-93. 5. Labar DR, et al. Neurology 2002;59:S38-43. 6. Labar DR. Seizure 2004;13:392-8. 7. Amar
AP, et al. Stereotact Funct Neurosurg 1999;73:104-8. 8.Ghaemi K et al. Seizure,2010;19:264-268(6.9%).
Slide courtesy of Sophia Varadkar

55. New Generation VNS Therapy
AspireSR What’s new? Cardiac-
based seizure detection
– (Standard VNS Therapy
stimulation with on-demand
magnet stimulation)
– Seizure detection
algorithm based on ictal
tachycardia
– Automatic stimulation
upon seizure detection

56. Seizure Detection Algorithm
• Senses cardiac R-waves
between the negative
electrode of lead and the
pulse generator
• Continuously analyses
changes in relative heart
rate
• If heart-rate goes above
programmable threshold
– % increase from baseline
– Automatic Stimulation is
triggered
VNS
Sensing
Vector
Slide courtesy of Sophia Varadkar

57. UCL - Institute of Child Health
Even Newer Generation VNS Therapy
SenTiva
Personalised features
– Guided programming
– Scheduled
programming
– Day and night
programming
Slide courtesy of Sophia Varadkar

58. Consider VNS Therapy if
• drug resistant epilepsy and resective surgery not an option
• multifocal seizures
• generalised seizures in epilepsy with bilateral structural/ metabolic or genetic
aetiology
• idiopathic generalised epilepsies

59. • Possible areas for deep brain stimulation
– Cerebellum
– Caudate
– Subthalamic Nucleus
– Substantia nigra (SNr)
– Thalamus
– Anterior Nucleus
– Centromedian Nucleus
DBS logic and objectives
Li and Cook Epilepsia 2017

60. • SANTE pivotal trial by Fisher et al 2010
• Double blind multicenter study
• 43% with ≥50% reduction in seizure
frequency at 1 year, and 68% at 5 years
Anterior thalamic stimulation

61. Centromedian nucleus stimulation
• Pioneered by Velasco in Mexico (1987)
• Closely related to the ascending reticular
activating system pathways and has
particularly central and frontal cortical
connections
• Appears to benefit bifrontal or generalised
seizures

62. • Status epilepticus in a localised cortical region:
Epilepsia partialis continua
– Chronic cortical stimulation if the area is localized
– 3 cases at Kings. EPC stopped in all cases.
DBS logic and objectives
Valentin et al, 2012

63. Cannibidiol or THC – all required?

64. Is there evidence for effect in epilepsy?
• Large preclinical evidence base asserting mixed effects on
seizures in animal models
64
Compound
Species
Number of discrete
conditions/models/designs
Dose Anticonvulsant No effect Proconvulsant
THC 6 31
0.25-200
mg/kg
61% 29% 10%*
CBD 2 21
1-400
mg/kg
81% 19% 0%
Other plant
cannabinoids
2 7 N/A 100% 0% 0%
CB1 receptor
agonists
2 55 N/A 73% 18%
2%
(7% mixed effect)
*Includes non-seizure studies where convulsions were reported (see next slide)
Whalley (2014) Cannabis and Seizures American Herbal Pharmacopeia

65. Cannabinoids: GW Pharma
• Pure cannabidiol and cannabidivarin; almost
insignificant THC
• CBD is one of two major cannabinoids in Sativex
• Epidiolex is a proprietary oral solution of pure
plant-derived cannabidiol

66. Cannabidiol in childhood epilepsy
• 214 patients across 11
sites
• safety & tolerability 167
• Adverse events 79%
• 137 efficacy analysis
Dravet N=32
49% responders, 3% SF
LGS N=30
37% responders, 3% SF
Lancet Neurology 2016;15:270-8

67. >50% reduction
• Clobazam 36/70 (51%)
• No clobazam 18/67 (27%)
Multiple logistic regression
clobazam use only
independent predictor of
reduction >50% in motor
seizures
Lancet Neurology 2016;15:270-8

68. Hemp oil
<0.2% THC
CBD 5%, THC 0.2%
Not the same!

69. >2% THC

70. • RCT data of short term efficacy of cannabidiol vs placebo in Dravet and
Lennox Gastaut syndrome
• Trials underway in Tuberous Sclerosis, Infantile Spasms
• Long term safety/tolerability & sustained efficacy requires evaluation
open label data suggests sustained efficacy
pure CBD vs ?THC No evidence as yet
• Possible wider benefits in comorbidities (eg anxiety disorders, cognition)
requires further consideration
• Read BPNA statement on cannabis
Cannabinoids in epilepsy
Evidence from clinical trials

71. Game changers
• Fenfluramine in Dravet syndrome
Polster Epilepsy and Behaviour 2018
• Anakinra (IL1Ra) in FIRES
Kenney-Jung et al Annals of Neurology 2016
• Gene therapy (e.g. CLN2 related Batten’s disease)
Donsante and Boulis Expert Opinion in Biological Therapy 2018

72. ALREADY HERE or NOT SO FAR AWAY

74. Epilepsy Team - Great Ormond Street Hospital
• Neurophysiology:
• Rachel Thornton
• Frideriecke Moeller
• Ronit Pressler
• Krishna Das
• Steward Boyd
• Neuropsychology:
• Prof Torsten Baldeweg
• Adam Kuczynski
• Adam Kuczynski
• Prof Faraneigh Vagha-Khan
• Neurodevelopmental Paediatrics:
• Maria Clark
• Hannah Richardson
• Neuropsychiatry:
• Prof Isobel Heyman
• Histopathology:
• Thomas Jaques
• Neurology:
• Sarah Aylett
• Prof Helen Cross
• Krishna Das
• Nivedita Desai
• Prof Finbar O’Callaghan
• Christin Eltze
• Amy McTague
• Robert Robinson
• Prof Rod Scott
• Sophia Varadkar
• Neurosurgery:
• Martin Tisdall
• Zubair Tahir
• Neuroradiology:
• Felice D’Arco
• Nuclear Medicine:
• Lorenzo Biassoni
Special thanks to Dr Christin Eltze, Prof Helen Cross, Dr Sophia Varadkar and
Dr Antonio Valentin for their permission to use some of their slides